mirror of
https://github.com/ROCm/composable_kernel.git
synced 2026-05-21 13:29:20 +00:00
b9f9ed96ac
* add DeviceGemmXdl
* update script
* fix naming issue
* fix comment
* output HostTensorDescriptor
* rename
* padded GEMM for fwd v4r4r4 nhwc
* refactor
* refactor
* refactor
* adding ckProfiler
* adding ckProfiler
* refactor
* fix tuning parameter bug
* add more gemm instances
* add more fp16 GEMM instances
* fix profiler driver
* fix bug in tuning parameter
* add fp32 gemm instances
* small fix
* refactor
* rename
* refactor gemm profiler; adding DeviceConv and conv profiler
* refactor
* fix
* add conv profiler
* refactor
* adding more GEMM and Conv instance
* Create README.md
Add build instruction for ckProfiler
* Create README.md
Add Readme for gemm_xdl example
* Update README.md
Remove build instruction from top most folder
* Update README.md
* clean up
[ROCm/composable_kernel commit: e823d518cb]
59 lines
2.0 KiB
C++
59 lines
2.0 KiB
C++
#ifndef DEVICE_CONV_FWD_XDL_HPP
|
|
#define DEVICE_CONV_FWD_XDL_HPP
|
|
|
|
#include <iostream>
|
|
#include "device.hpp"
|
|
#include "device_base.hpp"
|
|
#include "device_conv.hpp"
|
|
#include "common_header.hpp"
|
|
#include "tensor_layout.hpp"
|
|
#include "tensor_descriptor.hpp"
|
|
#include "tensor_descriptor_helper.hpp"
|
|
#include "gridwise_gemm_xdlops_v2r3.hpp"
|
|
|
|
namespace ck {
|
|
namespace tensor_operation {
|
|
namespace device {
|
|
|
|
template <ck::index_t NDimSpatial,
|
|
typename InDataType,
|
|
typename WeiDataType,
|
|
typename OutDataType,
|
|
typename AccDataType,
|
|
typename InLayout,
|
|
typename WeiLayout,
|
|
typename OutLayout,
|
|
ck::index_t BlockSize,
|
|
ck::index_t MPerBlock,
|
|
ck::index_t NPerBlock,
|
|
ck::index_t K0PerBlock,
|
|
ck::index_t K1,
|
|
ck::index_t MPerXDL,
|
|
ck::index_t NPerXDL,
|
|
ck::index_t MXdlPerWave,
|
|
ck::index_t NXdlPerWave,
|
|
typename ABlockTransferThreadSliceLengths_K0_M_K1,
|
|
typename ABlockTransferThreadClusterLengths_K0_M_K1,
|
|
typename ABlockTransferThreadClusterArrangeOrder,
|
|
typename ABlockTransferSrcAccessOrder,
|
|
ck::index_t ABlockTransferSrcVectorDim,
|
|
ck::index_t ABlockTransferSrcScalarPerVector,
|
|
ck::index_t ABlockTransferDstScalarPerVector_K1,
|
|
typename BBlockTransferThreadSliceLengths_K0_N_K1,
|
|
typename BBlockTransferThreadClusterLengths_K0_N_K1,
|
|
typename BBlockTransferThreadClusterArrangeOrder,
|
|
typename BBlockTransferSrcAccessOrder,
|
|
ck::index_t BBlockTransferSrcVectorDim,
|
|
ck::index_t BBlockTransferSrcScalarPerVector,
|
|
ck::index_t BBlockTransferDstScalarPerVector_K1,
|
|
ck::index_t CThreadTransferSrcDstVectorDim,
|
|
ck::index_t CThreadTransferDstScalarPerVector,
|
|
bool ABlockLdsAddExtraM,
|
|
bool BBlockLdsAddExtraN>
|
|
struct DeviceConvFwdXdl;
|
|
|
|
} // namespace device
|
|
} // namespace tensor_operation
|
|
} // namespace ck
|
|
#endif
|