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.

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

Minor performance optimization and bugfixes #233

Merged
merged 1 commit into from
Feb 23, 2018
Merged

Minor performance optimization and bugfixes #233

Show file tree
Hide file tree
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
2 changes: 1 addition & 1 deletion CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -89,7 +89,7 @@ option(${UPPER_PROJECT_NAME}_WITH_NVPROF
option(${UPPER_PROJECT_NAME}_WITH_TOPO_AWARE
"Enable topology-aware profiling (HWLOC)" ON)

option(${UPPER_PROJECT_NAME}_WITH_ALUNINUM
option(${UPPER_PROJECT_NAME}_WITH_ALUMINUM
"Enable Aluminum all-reduce library" OFF)

option(${UPPER_PROJECT_NAME}_VERBOSE "Enable verbose output" OFF)
Expand Down
1 change: 1 addition & 0 deletions src/utils/cudnn_wrapper.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -320,6 +320,7 @@ cudnn_manager::cudnn_manager(lbann::lbann_comm *_comm, int max_num_gpus, bool nc
FORCE_CHECK_CUDNN(cudnnCreate(&m_handles.back()));
FORCE_CHECK_CUDNN(cudnnSetStream(m_handles.back(), m_streams.back()));
FORCE_CHECK_CUBLAS(cublasCreate(&m_cublas_handles.back()));
FORCE_CHECK_CUBLAS(cublasSetStream(m_cublas_handles.back(), m_streams.back()));
}


Expand Down
2 changes: 1 addition & 1 deletion src/utils/cudnn_wrapper.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -147,7 +147,7 @@ void cudnn_manager::global_allreduce_on_gpus(std::vector<DataType*>& gpu_data,
El::mpi::Comm comm) {
if(!is_nccl_used()){
static Mat cpu_workspace;
cpu_workspace.Resize(height, width);
cpu_workspace.Resize(height, width, height);
allreduce_on_gpus(gpu_data, height, width);
copy_from_gpu(0, cpu_workspace, gpu_data[0]);
synchronize();
Expand Down
2 changes: 1 addition & 1 deletion superbuild/openblas/CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,7 +16,7 @@ set(OPENBLAS_CMAKE_INSTALL_PREFIX "${CMAKE_INSTALL_PREFIX}"
set(OPENBLAS_URL "https://github.com/xianyi/OpenBLAS.git"
CACHE STRING "The URL from which to clone OpenBLAS")

set(OPENBLAS_TAG "v0.2.15"
set(OPENBLAS_TAG "db72ad8"
CACHE STRING "The git tag or hash to checkout for OpenBLAS")

# Handle the options
Expand Down