4 ms·
"skipping Nvidia’s closed-source CUDA libraries, such as cuBLAS, in favor of open-source libraries, such as cutlass" Eww, so it'll be 90% slower? cuBLAS is in
by fxtentacle 4y ago
"skipping Nvidia’s closed-source CUDA libraries, such as cuBLAS, in favor of open-source libraries, such as cutlass"
Eww, so it'll be 90% slower?
cuBLAS is insanely optimized. It's the reason why CUDA dominates the field even though one could in theory use a ROCm build of TensorFlow or PyTorch. But performance is abysmal without cuBLAS. So the entire article appears to hinge on a huge assumption: What if Free Open Source was equivalent to NVIDIA's top-tier priorietary library? (It's not)
- cjbgkagh 4y agoCutlass is from NVidia and AFAIK gets pretty close to cuBLAS performance. I think the use case for cutlass is when you only need a few kernels and don’t want to pull in a huge cuBLAS dependency and are ok paying a small perf penalty for that. Then there are times when you need custom kernels that not available in cuBLAS and for that cutlass is about as fast as it gets.