DeepSeek/DeepGEMM
1
0
Fork 0
mirror of https://github.com/deepseek-ai/DeepGEMM synced 2025-06-26 23:15:49 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

Add DG_NVCC_OVERRIDE_CPP_STANDARD

Browse Source
This commit is contained in:
Chenggang Zhao
2025-04-03 15:53:29 +08:00
parent 3a5539b7db
commit d14962f072
2 changed files with 3 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
README.md
View File

@@ -128,6 +128,7 @@ The library also provides some environment variables, which may be useful:
- `DG_CACHE_DIR`: string, the cache directory to store compiled kernels, `$HOME/.deep_gemm` by default
- `DG_NVCC_COMPILER`: string, specified NVCC compiler path; will find in `from torch.utils.cpp_extension.CUDA_HOME` by default
- `DG_NVCC_OVERRIDE_CPP_STANDARD`: integer (e.g., `20`), support for some old version GCC compiler
- `DG_DISABLE_FFMA_INTERLEAVE`: 0 or 1, disable FFMA-interleaving optimization
- `DG_PTXAS_VERBOSE`: 0 or 1, show detailed PTXAS compiler output
- `DG_PRINT_REG_REUSE`: 0 or 1, print FFMA-interleaving details