Home / Class/ getWarpsPerSmFw Class — pytorch Architecture

getWarpsPerSmFw Class — pytorch Architecture

Architecture documentation for the getWarpsPerSmFw class in kernel_forward.h from the pytorch codebase.

Class c

Entity Profile

Source Code

aten/src/ATen/native/transformers/cuda/mem_eff_attention/kernel_forward.h lines 58–65

template <typename scalar_t, typename Arch>
constexpr int getWarpsPerSmFw() {
  return (
      Arch::kMinComputeCapability >= 80 &&
              !cutlass::platform::is_same<scalar_t, float>::value
          ? 16
          : 12);
}

Source

Analyze Your Own Codebase

Get architecture documentation, dependency graphs, and domain analysis for your codebase in minutes.

Try Supermodel Free