Skip to content

[Build] Reduce nvcc_threads to 1 for linux cuda plugin ci - #31729

Open
Tianlei Wu (tianleiwu) wants to merge 1 commit into
mainfrom
tlwu/cuda_plugin_ci_nvcc_threads_1
Open

[Build] Reduce nvcc_threads to 1 for linux cuda plugin ci#31729
Tianlei Wu (tianleiwu) wants to merge 1 commit into
mainfrom
tlwu/cuda_plugin_ci_nvcc_threads_1

Conversation

@tianleiwu

@tianleiwu Tianlei Wu (tianleiwu) commented Aug 8, 2026

Copy link
Copy Markdown
Contributor

Linux cuda plugin CI failed sometime due to OOM. This reduce nvcc_threads to 1 to avoid the issue.

Example failed jobs:
https://github.com/microsoft/onnxruntime/actions/runs/31229218069/job/93029489524?pr=31704

It seems that runner has no log after the following (looks like OOM):

[1133/1762] Building CUDA object CMakeFiles/onnxruntime_providers_cuda_plugin_sm90_tma.dir/onnxruntime_src/onnxruntime/contrib_ops/cuda/llm/fpA_intB_gemm/launchers/fpA_intB_gemm_launcher_7.generated.cu.o
[1134/1762] Building CUDA object CMakeFiles/onnxruntime_providers_cuda_plugin_sm90_tma.dir/onnxruntime_src/onnxruntime/contrib_ops/cuda/llm/fpA_intB_gemm/launchers/fpA_intB_gemm_launcher_8.generated.cu.o
[1135/1762] Building CUDA object CMakeFiles/onnxruntime_providers_cuda_plugin_sm90_tma.dir/onnxruntime_src/onnxruntime/contrib_ops/cuda/llm/fpA_intB_gemm/launchers/fpA_intB_gemm_launcher_9.generated.cu.o

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant