device_batched_gemm_wmma_cshuffle_v3.hpp File Reference

• include
• ck
• tensor_operation
• gpu
• device
• impl

Classes | Namespaces | Functions

device_batched_gemm_wmma_cshuffle_v3.hpp File Reference

#include <iostream>
#include <sstream>
#include "ck/utility/common_header.hpp"
#include "ck/tensor_description/tensor_descriptor.hpp"
#include "ck/tensor_description/tensor_descriptor_helper.hpp"
#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
#include "ck/tensor_operation/gpu/device/device_batched_gemm.hpp"
#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
#include "ck/tensor_operation/gpu/grid/gridwise_gemm_wmma_cshuffle_v3.hpp"
#include "ck/host_utility/device_prop.hpp"
#include "ck/host_utility/kernel_launch.hpp"
#include "ck/host_utility/flush_cache.hpp"

Go to the source code of this file.

Classes
struct	ck::tensor_operation::device::DeviceBatchedGemm_Wmma_CShuffleV3< ALayout, BLayout, CLayout, ADataType, BDataType, CDataType, AccDataType, CShuffleDataType, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, GemmSpec, BlockSize, MPerBlock, NPerBlock, KPerBlock, AK1, BK1, MPerWmma, NPerWmma, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_AK0_M_AK1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_AK1, ABlockLdsExtraM, BBlockTransferThreadClusterLengths_BK0_N_BK1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_BK1, BBlockLdsExtraN, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CShuffleBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, CShuffleBlockTransferScalarPerVector_NPerBlock, BlkGemmPipeSched, BlkGemmPipelineVer, ComputeTypeA, ComputeTypeB, PermuteA, PermuteB >
	"Universal" Batched GEMM operation without SplitK support. More...
struct	ck::tensor_operation::device::DeviceBatchedGemm_Wmma_CShuffleV3< ALayout, BLayout, CLayout, ADataType, BDataType, CDataType, AccDataType, CShuffleDataType, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, GemmSpec, BlockSize, MPerBlock, NPerBlock, KPerBlock, AK1, BK1, MPerWmma, NPerWmma, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_AK0_M_AK1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_AK1, ABlockLdsExtraM, BBlockTransferThreadClusterLengths_BK0_N_BK1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_BK1, BBlockLdsExtraN, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CShuffleBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, CShuffleBlockTransferScalarPerVector_NPerBlock, BlkGemmPipeSched, BlkGemmPipelineVer, ComputeTypeA, ComputeTypeB, PermuteA, PermuteB >::ComputePtrOffsetOfStridedBatch
struct	ck::tensor_operation::device::DeviceBatchedGemm_Wmma_CShuffleV3< ALayout, BLayout, CLayout, ADataType, BDataType, CDataType, AccDataType, CShuffleDataType, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, GemmSpec, BlockSize, MPerBlock, NPerBlock, KPerBlock, AK1, BK1, MPerWmma, NPerWmma, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_AK0_M_AK1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_AK1, ABlockLdsExtraM, BBlockTransferThreadClusterLengths_BK0_N_BK1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_BK1, BBlockLdsExtraN, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CShuffleBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, CShuffleBlockTransferScalarPerVector_NPerBlock, BlkGemmPipeSched, BlkGemmPipelineVer, ComputeTypeA, ComputeTypeB, PermuteA, PermuteB >::Argument
struct	ck::tensor_operation::device::DeviceBatchedGemm_Wmma_CShuffleV3< ALayout, BLayout, CLayout, ADataType, BDataType, CDataType, AccDataType, CShuffleDataType, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, GemmSpec, BlockSize, MPerBlock, NPerBlock, KPerBlock, AK1, BK1, MPerWmma, NPerWmma, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_AK0_M_AK1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_AK1, ABlockLdsExtraM, BBlockTransferThreadClusterLengths_BK0_N_BK1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_BK1, BBlockLdsExtraN, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CShuffleBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, CShuffleBlockTransferScalarPerVector_NPerBlock, BlkGemmPipeSched, BlkGemmPipelineVer, ComputeTypeA, ComputeTypeB, PermuteA, PermuteB >::Invoker
	Helper structure responsible for kernel invocation. More...

Namespaces
namespace	ck
namespace	ck::tensor_operation
namespace	ck::tensor_operation::device

Functions
template<typename GridwiseGemm, typename ComputePtrOffsetOfStridedBatch, bool HasMainKBlockLoop, InMemoryDataOperationEnum CGlobalMemoryDataOperation, index_t MinimumOccupancy = 1, TailNumber TailNum = TailNumber::Full>
__global__ void	ck::tensor_operation::device::kernel_batched_gemm_wmma_cshuffle_v3 (typename GridwiseGemm::Argument karg, const ComputePtrOffsetOfStridedBatch compute_ptr_offset_of_batch)