diff options
| author |  Gunhan Gulsoy <gunan@google.com> | 2018-07-24 11:10:46 -0700 |
|---|---|---|
| committer |  TensorFlower Gardener <gardener@tensorflow.org> | 2018-07-24 11:21:11 -0700 |
| commit | aef000ed3c2863a5cc7ccb5bf1fb46116e7f4f02 (patch) | |
| tree | 138ee7ae7f12f3b6bd617fb13a9cdfeacb860313 /tensorflow/contrib/cmake | |
| parent | 0527ba2f447fe0bc20152f393bcd672de0b59548 (diff) | |
Build more cuda compute capabilities in cmake build.
Fixes #18652
PiperOrigin-RevId: 205858348
Diffstat (limited to 'tensorflow/contrib/cmake')
| -rw-r--r-- | tensorflow/contrib/cmake/CMakeLists.txt | 14 |
1 files changed, 9 insertions, 5 deletions
diff --git a/tensorflow/contrib/cmake/CMakeLists.txt b/tensorflow/contrib/cmake/CMakeLists.txt index 708618dcb0..6c93487e0d 100644 --- a/tensorflow/contrib/cmake/CMakeLists.txt +++ b/tensorflow/contrib/cmake/CMakeLists.txt @@ -394,16 +394,20 @@ if (tensorflow_ENABLE_GPU) # by default we assume compute cabability 3.5 and 5.2. If you change this change it in # CUDA_NVCC_FLAGS and cuda_config.h below - set(CUDA_NVCC_FLAGS ${CUDA_NVCC_FLAGS};-gencode arch=compute_30,code=\"sm_30,compute_30\";-gencode arch=compute_35,code=\"sm_35,compute_35\";-gencode arch=compute_52,code=\"sm_52,compute_52\") + set(CUDA_NVCC_FLAGS ${CUDA_NVCC_FLAGS};-gencode arch=compute_37,code=\"sm_37,compute_37\") + set(CUDA_NVCC_FLAGS ${CUDA_NVCC_FLAGS};-gencode arch=compute_52,code=\"sm_52,compute_52\") + set(CUDA_NVCC_FLAGS ${CUDA_NVCC_FLAGS};-gencode arch=compute_60,code=\"sm_60,compute_60\") + set(CUDA_NVCC_FLAGS ${CUDA_NVCC_FLAGS};-gencode arch=compute_61,code=\"sm_61,compute_61\") + set(CUDA_NVCC_FLAGS ${CUDA_NVCC_FLAGS};-gencode arch=compute_70,code=\"sm_70,compute_70\") set(CUDA_NVCC_FLAGS ${CUDA_NVCC_FLAGS};--include-path ${PROJECT_BINARY_DIR}/$\{build_configuration\};--expt-relaxed-constexpr) set(CUDA_NVCC_FLAGS ${CUDA_NVCC_FLAGS};-ftz=true) # Flush denormals to zero set(CUDA_INCLUDE ${CUDA_TOOLKIT_TARGET_DIR} ${CUDA_TOOLKIT_TARGET_DIR}/extras/CUPTI/include) include_directories(${CUDA_INCLUDE}) if (WIN32) - add_definitions(-DGOOGLE_CUDA=1 -DTF_EXTRA_CUDA_CAPABILITIES=3.0,3.5,5.2) + add_definitions(-DGOOGLE_CUDA=1 -DTF_EXTRA_CUDA_CAPABILITIES=3.7,5.2,6.0,6.1,7.0) else (WIN32) - # Without these double quotes, cmake in Linux makes it "-DTF_EXTRA_CUDA_CAPABILITIES=3.0, -D3.5, -D5.2" for cc, which incurs build breaks - add_definitions(-DGOOGLE_CUDA=1 -D"TF_EXTRA_CUDA_CAPABILITIES=3.0,3.5,5.2") + # Without these double quotes, cmake in Linux makes it "-DTF_EXTRA_CUDA_CAPABILITIES=3.7, -D5.2, ..." for cc, which incurs build breaks + add_definitions(-DGOOGLE_CUDA=1 -D"TF_EXTRA_CUDA_CAPABILITIES=3.7,5.2,6.0,6.1,7.0") endif (WIN32) if (WIN32) @@ -452,7 +456,7 @@ if (tensorflow_ENABLE_GPU) FILE(WRITE ${tensorflow_source_dir}/third_party/gpus/cuda/cuda_config.h "#ifndef CUDA_CUDA_CONFIG_H_\n" "#define CUDA_CUDA_CONFIG_H_\n" - "#define TF_CUDA_CAPABILITIES CudaVersion(\"3.0\"),CudaVersion(\"3.5\"),CudaVersion(\"5.2\")\n" + "#define TF_CUDA_CAPABILITIES CudaVersion(\"3.7\"),CudaVersion(\"5.2\"),CudaVersion(\"6.0\"),CudaVersion(\"6.1\"),CudaVersion(\"7.0\")\n" "#define TF_CUDA_VERSION \"64_${short_CUDA_VER}\"\n" "#define TF_CUDNN_VERSION \"64_${tensorflow_CUDNN_VERSION}\"\n" "#define TF_CUDA_TOOLKIT_PATH \"${CUDA_TOOLKIT_ROOT_DIR}\"\n" |