initial cuda run

[ROCm/composable_kernel commit: 49ceb0fe04]
This commit is contained in:
Chao Liu
2018-10-22 20:51:26 -05:00
parent 6a5c465ad9
commit 0de357304a
2 changed files with 15 additions and 17 deletions

View File

@@ -3,10 +3,10 @@
template <class TFloat, int NBlockDim>
__global__ void direct_convolution(DeviceTensorDescriptor in_desc,
TFloat* const in,
TFloat* const p_in,
DeviceTensorDescriptor wei_desc,
TFloat* const wei,
TFloat* const p_wei,
DeviceTensorDescriptor out_desc,
TFloat* out)
TFloat* p_out)
{
}