Repository navigation

CUDA backend: optional cuBLAS for matrix multiplications - #24129

Closed

shoumikhin wants to merge 2 commits into

pytorch:mainfrom

shoumikhin:cuda-cublas-gemm

Closed

CUDA backend: optional cuBLAS for matrix multiplications#24129
shoumikhin wants to merge 2 commits into
pytorch:mainfrom
shoumikhin:cuda-cublas-gemm

Commits on Oct 24, 2026
- CUDA backend: add cuBLAS shims for mm, addmm and bmm
  shoumikhin
  committed
  1832491
- CUDA backend: let autotuning pick cuBLAS for matrix multiplications
  shoumikhin
  committed
  3865daa