Composable Kernel: ck::tensor_operation::device::DeviceGemmDl< ADataType, BDataType, CDataType, AccDataType, ALayout, BLayout, CLayout, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, GemmSpec, BlockSize, MPerBlock, NPerBlock, K0PerBlock, K1, M1PerThread, N1PerThread, KPerThread, M1N1ThreadClusterM1Xs, M1N1ThreadClusterN1Xs, ABlockTransferThreadSliceLengths_K0_M0_M1_K1, ABlockTransferThreadClusterLengths_K0_M0_M1_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorTensorLengths_K0_M0_M1_K1, ABlockTransferSrcVectorTensorContiguousDimOrder, ABlockTransferDstVectorTensorLengths_K0_M0_M1_K1, BBlockTransferThreadSliceLengths_K0_N0_N1_K1, BBlockTransferThreadClusterLengths_K0_N0_N1_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorTensorLengths_K0_N0_N1_K1, BBlockTransferSrcVectorTensorContiguousDimOrder, BBlockTransferDstVectorTensorLengths_K0_N0_N1_K1, CThreadTransferSrcDstAccessOrder, CThreadTransferSrcDstVectorDim, CThreadTransferDstScalarPerVector, >::Invoker Struct Reference

Invoker Struct Reference

Invoker Struct Reference#

Composable Kernel: ck::tensor_operation::device::DeviceGemmDl< ADataType, BDataType, CDataType, AccDataType, ALayout, BLayout, CLayout, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, GemmSpec, BlockSize, MPerBlock, NPerBlock, K0PerBlock, K1, M1PerThread, N1PerThread, KPerThread, M1N1ThreadClusterM1Xs, M1N1ThreadClusterN1Xs, ABlockTransferThreadSliceLengths_K0_M0_M1_K1, ABlockTransferThreadClusterLengths_K0_M0_M1_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorTensorLengths_K0_M0_M1_K1, ABlockTransferSrcVectorTensorContiguousDimOrder, ABlockTransferDstVectorTensorLengths_K0_M0_M1_K1, BBlockTransferThreadSliceLengths_K0_N0_N1_K1, BBlockTransferThreadClusterLengths_K0_N0_N1_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorTensorLengths_K0_N0_N1_K1, BBlockTransferSrcVectorTensorContiguousDimOrder, BBlockTransferDstVectorTensorLengths_K0_N0_N1_K1, CThreadTransferSrcDstAccessOrder, CThreadTransferSrcDstVectorDim, CThreadTransferDstScalarPerVector, >::Invoker Struct Reference

Public Types | Public Member Functions | List of all members

ck::tensor_operation::device::DeviceGemmDl< ADataType, BDataType, CDataType, AccDataType, ALayout, BLayout, CLayout, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, GemmSpec, BlockSize, MPerBlock, NPerBlock, K0PerBlock, K1, M1PerThread, N1PerThread, KPerThread, M1N1ThreadClusterM1Xs, M1N1ThreadClusterN1Xs, ABlockTransferThreadSliceLengths_K0_M0_M1_K1, ABlockTransferThreadClusterLengths_K0_M0_M1_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorTensorLengths_K0_M0_M1_K1, ABlockTransferSrcVectorTensorContiguousDimOrder, ABlockTransferDstVectorTensorLengths_K0_M0_M1_K1, BBlockTransferThreadSliceLengths_K0_N0_N1_K1, BBlockTransferThreadClusterLengths_K0_N0_N1_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorTensorLengths_K0_N0_N1_K1, BBlockTransferSrcVectorTensorContiguousDimOrder, BBlockTransferDstVectorTensorLengths_K0_N0_N1_K1, CThreadTransferSrcDstAccessOrder, CThreadTransferSrcDstVectorDim, CThreadTransferDstScalarPerVector, >::Invoker Struct Reference

#include <device_gemm_dl.hpp>

Inheritance diagram for ck::tensor_operation::device::DeviceGemmDl< ADataType, BDataType, CDataType, AccDataType, ALayout, BLayout, CLayout, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, GemmSpec, BlockSize, MPerBlock, NPerBlock, K0PerBlock, K1, M1PerThread, N1PerThread, KPerThread, M1N1ThreadClusterM1Xs, M1N1ThreadClusterN1Xs, ABlockTransferThreadSliceLengths_K0_M0_M1_K1, ABlockTransferThreadClusterLengths_K0_M0_M1_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorTensorLengths_K0_M0_M1_K1, ABlockTransferSrcVectorTensorContiguousDimOrder, ABlockTransferDstVectorTensorLengths_K0_M0_M1_K1, BBlockTransferThreadSliceLengths_K0_N0_N1_K1, BBlockTransferThreadClusterLengths_K0_N0_N1_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorTensorLengths_K0_N0_N1_K1, BBlockTransferSrcVectorTensorContiguousDimOrder, BBlockTransferDstVectorTensorLengths_K0_N0_N1_K1, CThreadTransferSrcDstAccessOrder, CThreadTransferSrcDstVectorDim, CThreadTransferDstScalarPerVector, >::Invoker:

Public Types
using	Argument = DeviceGemmDl::Argument

Public Member Functions
float	Run (const Argument &arg, const StreamConfig &stream_config=StreamConfig{})
float	Run (const BaseArgument *p_arg, const StreamConfig &stream_config=StreamConfig{}) override
Public Member Functions inherited from ck::tensor_operation::device::BaseInvoker
	BaseInvoker ()=default
	BaseInvoker (const BaseInvoker &)=default
BaseInvoker &	operator= (const BaseInvoker &)=default
virtual	~BaseInvoker ()

Member Typedef Documentation

◆ Argument

template<typename ADataType, typename BDataType, typename CDataType, typename AccDataType, typename ALayout, typename BLayout, typename CLayout, typename AElementwiseOperation, typename BElementwiseOperation, typename CElementwiseOperation, GemmSpecialization GemmSpec, index_t BlockSize, index_t MPerBlock, index_t NPerBlock, index_t K0PerBlock, index_t K1, index_t M1PerThread, index_t N1PerThread, index_t KPerThread, typename M1N1ThreadClusterM1Xs, typename M1N1ThreadClusterN1Xs, typename ABlockTransferThreadSliceLengths_K0_M0_M1_K1, typename ABlockTransferThreadClusterLengths_K0_M0_M1_K1, typename ABlockTransferThreadClusterArrangeOrder, typename ABlockTransferSrcAccessOrder, typename ABlockTransferSrcVectorTensorLengths_K0_M0_M1_K1, typename ABlockTransferSrcVectorTensorContiguousDimOrder, typename ABlockTransferDstVectorTensorLengths_K0_M0_M1_K1, typename BBlockTransferThreadSliceLengths_K0_N0_N1_K1, typename BBlockTransferThreadClusterLengths_K0_N0_N1_K1, typename BBlockTransferThreadClusterArrangeOrder, typename BBlockTransferSrcAccessOrder, typename BBlockTransferSrcVectorTensorLengths_K0_N0_N1_K1, typename BBlockTransferSrcVectorTensorContiguousDimOrder, typename BBlockTransferDstVectorTensorLengths_K0_N0_N1_K1, typename CThreadTransferSrcDstAccessOrder, index_t CThreadTransferSrcDstVectorDim, index_t CThreadTransferDstScalarPerVector, enable_if_t< is_same_v< AElementwiseOperation, ck::tensor_operation::element_wise::PassThrough > &&is_same_v< BElementwiseOperation, ck::tensor_operation::element_wise::PassThrough > &&is_same_v< CElementwiseOperation, ck::tensor_operation::element_wise::PassThrough >, bool > = false>

using ck::tensor_operation::device::DeviceGemmDl< ADataType, BDataType, CDataType, AccDataType, ALayout, BLayout, CLayout, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, GemmSpec, BlockSize, MPerBlock, NPerBlock, K0PerBlock, K1, M1PerThread, N1PerThread, KPerThread, M1N1ThreadClusterM1Xs, M1N1ThreadClusterN1Xs, ABlockTransferThreadSliceLengths_K0_M0_M1_K1, ABlockTransferThreadClusterLengths_K0_M0_M1_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorTensorLengths_K0_M0_M1_K1, ABlockTransferSrcVectorTensorContiguousDimOrder, ABlockTransferDstVectorTensorLengths_K0_M0_M1_K1, BBlockTransferThreadSliceLengths_K0_N0_N1_K1, BBlockTransferThreadClusterLengths_K0_N0_N1_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorTensorLengths_K0_N0_N1_K1, BBlockTransferSrcVectorTensorContiguousDimOrder, BBlockTransferDstVectorTensorLengths_K0_N0_N1_K1, CThreadTransferSrcDstAccessOrder, CThreadTransferSrcDstVectorDim, CThreadTransferDstScalarPerVector, >::Invoker::Argument = DeviceGemmDl::Argument

Member Function Documentation

◆ Run() [1/2]

template<typename ADataType, typename BDataType, typename CDataType, typename AccDataType, typename ALayout, typename BLayout, typename CLayout, typename AElementwiseOperation, typename BElementwiseOperation, typename CElementwiseOperation, GemmSpecialization GemmSpec, index_t BlockSize, index_t MPerBlock, index_t NPerBlock, index_t K0PerBlock, index_t K1, index_t M1PerThread, index_t N1PerThread, index_t KPerThread, typename M1N1ThreadClusterM1Xs, typename M1N1ThreadClusterN1Xs, typename ABlockTransferThreadSliceLengths_K0_M0_M1_K1, typename ABlockTransferThreadClusterLengths_K0_M0_M1_K1, typename ABlockTransferThreadClusterArrangeOrder, typename ABlockTransferSrcAccessOrder, typename ABlockTransferSrcVectorTensorLengths_K0_M0_M1_K1, typename ABlockTransferSrcVectorTensorContiguousDimOrder, typename ABlockTransferDstVectorTensorLengths_K0_M0_M1_K1, typename BBlockTransferThreadSliceLengths_K0_N0_N1_K1, typename BBlockTransferThreadClusterLengths_K0_N0_N1_K1, typename BBlockTransferThreadClusterArrangeOrder, typename BBlockTransferSrcAccessOrder, typename BBlockTransferSrcVectorTensorLengths_K0_N0_N1_K1, typename BBlockTransferSrcVectorTensorContiguousDimOrder, typename BBlockTransferDstVectorTensorLengths_K0_N0_N1_K1, typename CThreadTransferSrcDstAccessOrder, index_t CThreadTransferSrcDstVectorDim, index_t CThreadTransferDstScalarPerVector, enable_if_t< is_same_v< AElementwiseOperation, ck::tensor_operation::element_wise::PassThrough > &&is_same_v< BElementwiseOperation, ck::tensor_operation::element_wise::PassThrough > &&is_same_v< CElementwiseOperation, ck::tensor_operation::element_wise::PassThrough >, bool > = false>

float ck::tensor_operation::device::DeviceGemmDl< ADataType, BDataType, CDataType, AccDataType, ALayout, BLayout, CLayout, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, GemmSpec, BlockSize, MPerBlock, NPerBlock, K0PerBlock, K1, M1PerThread, N1PerThread, KPerThread, M1N1ThreadClusterM1Xs, M1N1ThreadClusterN1Xs, ABlockTransferThreadSliceLengths_K0_M0_M1_K1, ABlockTransferThreadClusterLengths_K0_M0_M1_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorTensorLengths_K0_M0_M1_K1, ABlockTransferSrcVectorTensorContiguousDimOrder, ABlockTransferDstVectorTensorLengths_K0_M0_M1_K1, BBlockTransferThreadSliceLengths_K0_N0_N1_K1, BBlockTransferThreadClusterLengths_K0_N0_N1_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorTensorLengths_K0_N0_N1_K1, BBlockTransferSrcVectorTensorContiguousDimOrder, BBlockTransferDstVectorTensorLengths_K0_N0_N1_K1, CThreadTransferSrcDstAccessOrder, CThreadTransferSrcDstVectorDim, CThreadTransferDstScalarPerVector, >::Invoker::Run	(	const Argument &	arg,
		const StreamConfig &	stream_config = StreamConfig{} )

inline

◆ Run() [2/2]

template<typename ADataType, typename BDataType, typename CDataType, typename AccDataType, typename ALayout, typename BLayout, typename CLayout, typename AElementwiseOperation, typename BElementwiseOperation, typename CElementwiseOperation, GemmSpecialization GemmSpec, index_t BlockSize, index_t MPerBlock, index_t NPerBlock, index_t K0PerBlock, index_t K1, index_t M1PerThread, index_t N1PerThread, index_t KPerThread, typename M1N1ThreadClusterM1Xs, typename M1N1ThreadClusterN1Xs, typename ABlockTransferThreadSliceLengths_K0_M0_M1_K1, typename ABlockTransferThreadClusterLengths_K0_M0_M1_K1, typename ABlockTransferThreadClusterArrangeOrder, typename ABlockTransferSrcAccessOrder, typename ABlockTransferSrcVectorTensorLengths_K0_M0_M1_K1, typename ABlockTransferSrcVectorTensorContiguousDimOrder, typename ABlockTransferDstVectorTensorLengths_K0_M0_M1_K1, typename BBlockTransferThreadSliceLengths_K0_N0_N1_K1, typename BBlockTransferThreadClusterLengths_K0_N0_N1_K1, typename BBlockTransferThreadClusterArrangeOrder, typename BBlockTransferSrcAccessOrder, typename BBlockTransferSrcVectorTensorLengths_K0_N0_N1_K1, typename BBlockTransferSrcVectorTensorContiguousDimOrder, typename BBlockTransferDstVectorTensorLengths_K0_N0_N1_K1, typename CThreadTransferSrcDstAccessOrder, index_t CThreadTransferSrcDstVectorDim, index_t CThreadTransferDstScalarPerVector, enable_if_t< is_same_v< AElementwiseOperation, ck::tensor_operation::element_wise::PassThrough > &&is_same_v< BElementwiseOperation, ck::tensor_operation::element_wise::PassThrough > &&is_same_v< CElementwiseOperation, ck::tensor_operation::element_wise::PassThrough >, bool > = false>

float ck::tensor_operation::device::DeviceGemmDl< ADataType, BDataType, CDataType, AccDataType, ALayout, BLayout, CLayout, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, GemmSpec, BlockSize, MPerBlock, NPerBlock, K0PerBlock, K1, M1PerThread, N1PerThread, KPerThread, M1N1ThreadClusterM1Xs, M1N1ThreadClusterN1Xs, ABlockTransferThreadSliceLengths_K0_M0_M1_K1, ABlockTransferThreadClusterLengths_K0_M0_M1_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorTensorLengths_K0_M0_M1_K1, ABlockTransferSrcVectorTensorContiguousDimOrder, ABlockTransferDstVectorTensorLengths_K0_M0_M1_K1, BBlockTransferThreadSliceLengths_K0_N0_N1_K1, BBlockTransferThreadClusterLengths_K0_N0_N1_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorTensorLengths_K0_N0_N1_K1, BBlockTransferSrcVectorTensorContiguousDimOrder, BBlockTransferDstVectorTensorLengths_K0_N0_N1_K1, CThreadTransferSrcDstAccessOrder, CThreadTransferSrcDstVectorDim, CThreadTransferDstScalarPerVector, >::Invoker::Run	(	const BaseArgument *	p_arg,
		const StreamConfig &	stream_config = StreamConfig{} )

inlineoverridevirtual

Reimplemented from ck::tensor_operation::device::BaseInvoker.

The documentation for this struct was generated from the following file:

device_gemm_dl.hpp