zhangyiss/mlx
1
0
Fork 0
mirror of https://github.com/ml-explore/mlx.git synced 2025-12-16 01:49:05 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

use arch specific targets when possible (#2771)

Browse Source
This commit is contained in:
Awni Hannun
2025-11-14 20:04:18 -08:00
committed by GitHub
parent 3c622ddd1d
commit 1bf605d56d
4 changed files with 34 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
mlx/backend/cuda/CMakeLists.txt
View File

@@ -126,7 +126,11 @@ endif()
# Compute capability >= 7.0 is required for synchronization between CPU/GPU with
# managed memory.
if(NOT DEFINED MLX_CUDA_ARCHITECTURES)
set(MLX_CUDA_ARCHITECTURES "native")
execute_process(
COMMAND bash detect_cuda_arch.sh
WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}
OUTPUT_VARIABLE MLX_CUDA_ARCHITECTURES
OUTPUT_STRIP_TRAILING_WHITESPACE)
endif()
message(STATUS "CUDA architectures: ${MLX_CUDA_ARCHITECTURES}")
set_target_properties(mlx PROPERTIES CUDA_ARCHITECTURES