devtools/eigen
1
0
Fork 0
mirror of https://gitlab.com/libeigen/eigen.git synced 2026-04-10 11:34:33 +08:00
Code Issues Packages Projects Releases Wiki Activity

Added support for multi gpu configuration to the GpuDevice class

Browse Source
This commit is contained in:
Benoit Steiner
2015-07-15 12:38:34 -07:00
parent f5aa640862
commit e892524efe
3 changed files with 199 additions and 58 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
unsupported/Eigen/CXX11/src/Tensor/TensorConvolution.h
View File

@@ -835,10 +835,10 @@ struct TensorEvaluator<const TensorConvolutionOp<Indices, InputArgType, KernelAr
void executeEval(Scalar* data) const {
typedef typename TensorEvaluator<InputArgType, GpuDevice>::Dimensions InputDims;
const int maxSharedMem = sharedMemPerBlock();
const int maxThreadsPerBlock = maxCudaThreadsPerBlock();
const int maxBlocksPerProcessor = maxCudaThreadsPerMultiProcessor() / maxThreadsPerBlock;
const int numMultiProcessors = getNumCudaMultiProcessors();
const int maxSharedMem = m_device.sharedMemPerBlock();
const int maxThreadsPerBlock = m_device.maxCudaThreadsPerBlock();
const int maxBlocksPerProcessor = m_device.maxCudaThreadsPerMultiProcessor() / maxThreadsPerBlock;
const int numMultiProcessors = m_device.getNumCudaMultiProcessors();
const int warpSize = 32;
switch (NumKernelDims) {