22:39
2026-10-08
yang-yifan.github.io
ai-infrastructure
How to Make a Compute-Bound Problem Actually Compute-Bound
Anthropic member of technical staff Yifan Yang reported that a CUDA-core fp32 GEMM kernel on an Nvidia V100 GPU reached 10.7 TFLOPS, or 88% of cuBLAS's 12.1 TFLOPS, using only two optimizations over a…