Open chopin1998 opened 1 month ago
`❯ ./vkpeak 2 device = NVIDIA GeForce RTX 2080 Ti
fp32-scalar = 18017.38 GFLOPS fp32-vec4 = 17878.88 GFLOPS
fp16-scalar = 17801.07 GFLOPS fp16-vec4 = 19883.39 GFLOPS fp16-matrix = 122191.71 GFLOPS
fp64-scalar = 553.94 GFLOPS fp64-vec4 = 553.91 GFLOPS
int32-scalar = 16733.99 GIOPS int32-vec4 = 16466.83 GIOPS
int16-scalar = 10675.61 GIOPS int16-vec4 = 11735.68 GIOPS `
`❯ ./vkpeak 0 device = Intel(R) Graphics (RPL-S)
fp32-scalar = 686.73 GFLOPS fp32-vec4 = 608.83 GFLOPS
fp16-scalar = 919.97 GFLOPS fp16-vec4 = 1305.24 GFLOPS fp16-matrix = 148.76 GFLOPS
fp64-scalar = 0.00 GFLOPS fp64-vec4 = 0.00 GFLOPS
int32-scalar = 174.10 GIOPS int32-vec4 = 200.75 GIOPS
int16-scalar = 371.72 GIOPS int16-vec4 = 706.49 GIOPS `
FYI
`❯ ./vkpeak 2 device = NVIDIA GeForce RTX 2080 Ti
fp32-scalar = 18017.38 GFLOPS fp32-vec4 = 17878.88 GFLOPS
fp16-scalar = 17801.07 GFLOPS fp16-vec4 = 19883.39 GFLOPS fp16-matrix = 122191.71 GFLOPS
fp64-scalar = 553.94 GFLOPS fp64-vec4 = 553.91 GFLOPS
int32-scalar = 16733.99 GIOPS int32-vec4 = 16466.83 GIOPS
int16-scalar = 10675.61 GIOPS int16-vec4 = 11735.68 GIOPS `
`❯ ./vkpeak 0 device = Intel(R) Graphics (RPL-S)
fp32-scalar = 686.73 GFLOPS fp32-vec4 = 608.83 GFLOPS
fp16-scalar = 919.97 GFLOPS fp16-vec4 = 1305.24 GFLOPS fp16-matrix = 148.76 GFLOPS
fp64-scalar = 0.00 GFLOPS fp64-vec4 = 0.00 GFLOPS
int32-scalar = 174.10 GIOPS int32-vec4 = 200.75 GIOPS
int16-scalar = 371.72 GIOPS int16-vec4 = 706.49 GIOPS `
FYI