Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Test build time for sm 8.0 only #3068

Open
wants to merge 2 commits into
base: main
Choose a base branch
from
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion .github/scripts/fbgemm_gpu_build.bash
Original file line number Diff line number Diff line change
Expand Up @@ -199,7 +199,7 @@ __configure_fbgemm_gpu_build_cuda () {
# https://github.com/NVIDIA/nvbench/discussions/129
# https://github.com/vllm-project/vllm/blob/main/CMakeLists.txt#L187
# https://github.com/NVIDIA/cutlass/blob/main/include/cutlass/gemm/kernel/sm90_gemm_tma_warpspecialized.hpp#L224
local arch_list="7.0;8.0;9.0;9.0a"
local arch_list="7.0;8.0;9.0a"
else
local arch_list="7.0;8.0"
fi
Expand Down
4 changes: 2 additions & 2 deletions .github/scripts/nova_dir.bash
Original file line number Diff line number Diff line change
Expand Up @@ -16,5 +16,5 @@ export BUILD_FROM_NOVA=1
## Overwrite existing ENV VAR in Nova
if [[ "$CONDA_ENV" != "" ]]; then export CONDA_RUN="conda run --no-capture-output -p ${CONDA_ENV}" && echo "$CONDA_RUN"; fi
if [[ "$CU_VERSION" == "cu118" ]]; then export TORCH_CUDA_ARCH_LIST='7.0;8.0' && echo "$TORCH_CUDA_ARCH_LIST"; fi
if [[ "$CU_VERSION" == "cu121" ]]; then export TORCH_CUDA_ARCH_LIST='7.0;8.0;9.0;9.0a' && echo "$TORCH_CUDA_ARCH_LIST"; fi
if [[ "$CU_VERSION" == "cu124" ]]; then export TORCH_CUDA_ARCH_LIST='7.0;8.0;9.0;9.0a' && echo "$TORCH_CUDA_ARCH_LIST"; fi
if [[ "$CU_VERSION" == "cu121" ]]; then export TORCH_CUDA_ARCH_LIST='7.0;8.0;9.0a' && echo "$TORCH_CUDA_ARCH_LIST"; fi
if [[ "$CU_VERSION" == "cu124" ]]; then export TORCH_CUDA_ARCH_LIST='7.0;8.0;9.0a' && echo "$TORCH_CUDA_ARCH_LIST"; fi
1 change: 1 addition & 0 deletions CMakeLists.txt
Original file line number Diff line number Diff line change
Expand Up @@ -203,6 +203,7 @@ else(MSVC)
string(APPEND CMAKE_CXX_FLAGS " -Wno-deprecated-declarations")
string(APPEND CMAKE_CXX_FLAGS " -Wunknown-pragmas")
string(APPEND CMAKE_CXX_FLAGS " -Wimplicit-fallthrough")
string(APPEND CMAKE_CXX_FLAGS " -Wno-strict-aliasing")
target_compile_options(fbgemm_avx2 PRIVATE
"-m64" "-mavx2" "-mf16c" "-mfma")
target_compile_options(fbgemm_avx512 PRIVATE
Expand Down
6 changes: 3 additions & 3 deletions src/EmbeddingSpMDMAutovec.cc
Original file line number Diff line number Diff line change
Expand Up @@ -571,8 +571,8 @@ bool EmbeddingSpMDM_autovec(
return false;
}

constexpr int tile_size = 4;
#if _OPENMP >= 202011
constexpr int tile_size = 4;
#pragma omp tile sizes(tile_size)
#endif
for (int i = 0; i < len; ++i) {
Expand Down Expand Up @@ -711,8 +711,8 @@ bool EmbeddingSpMDMRowWiseSparse_autovec(
return false;
}

constexpr int tile_size = 4;
#if _OPENMP >= 202011
constexpr int tile_size = 4;
#pragma omp tile sizes(tile_size)
#endif
for (int i = 0; i < len; ++i) {
Expand Down Expand Up @@ -911,8 +911,8 @@ bool EmbeddingSpMDMFP8_autovec(
return false;
}

constexpr int tile_size = 4;
#if _OPENMP >= 202011
constexpr int tile_size = 4;
#pragma omp tile sizes(tile_size)
#endif
for (int i = 0; i < len; ++i) {
Expand Down
Loading