Skip to content

[src] Make cublas and cusparse use per-thread streams. #2692

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.

Sign up for GitHub

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

Merged
danpovey merged 1 commit into from
Sep 12, 2018
Merged

[src] Make cublas and cusparse use per-thread streams. #2692

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
5 changes: 4 additions & 1 deletion src/cudamatrix/cu-device.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -109,9 +109,10 @@ void CuDevice::Initialize() {
cudaSetDevice(device_id_);
// Initialize CUBLAS.
CUBLAS_SAFE_CALL(cublasCreate(&cublas_handle_));
CUBLAS_SAFE_CALL(cublasSetStream(cublas_handle_, cudaStreamPerThread));
Copy link
Contributor

@danpovey danpovey Sep 12, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@galv, thanks. I believe justin mentioned that we also need to use this cudaStreamPerThread for other API calls, like memcpy.

// Initialize the cuSPARSE library
CUSPARSE_SAFE_CALL(cusparseCreate(&cusparse_handle_));

CUSPARSE_SAFE_CALL(cusparseSetStream(cusparse_handle_, cudaStreamPerThread));
}
}

Expand Down Expand Up @@ -243,8 +244,10 @@ void CuDevice::FinalizeActiveGpu() {
// the main thread.
// Initialize CUBLAS.
CUBLAS_SAFE_CALL(cublasCreate(&cublas_handle_));
CUBLAS_SAFE_CALL(cublasSetStream(cublas_handle_, cudaStreamPerThread));
// Initialize the cuSPARSE library
CUSPARSE_SAFE_CALL(cusparseCreate(&cusparse_handle_));
CUSPARSE_SAFE_CALL(cusparseSetStream(cusparse_handle_, cudaStreamPerThread));

// Notify the user which GPU is being userd.
char name[128];
Expand Down