gridwise_gemm_xdlops_v2r4.hpp File Reference

• include
• ck
• tensor_operation
• gpu
• grid

Classes | Namespaces | Functions

gridwise_gemm_xdlops_v2r4.hpp File Reference

#include "ck/utility/common_header.hpp"
#include "ck/tensor_description/multi_index_transform_helper.hpp"
#include "ck/tensor_description/tensor_descriptor.hpp"
#include "ck/tensor_description/tensor_descriptor_helper.hpp"
#include "ck/tensor_operation/gpu/grid/block_to_ctile_map.hpp"
#include "ck/tensor_operation/gpu/grid/gridwise_gemm_pipeline_v1.hpp"
#include "ck/tensor_operation/gpu/block/blockwise_gemm_xdlops.hpp"
#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v4r1.hpp"
#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"
#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"

Go to the source code of this file.

Classes

struct

ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4< BlockSize, FloatAB, FloatAcc, FloatC, CGlobalMemoryDataOperation, ABK0MK1GridDesc, BBK0NK1GridDesc, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXDL, NPerXDL, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, CThreadTransferSrcDstAccessOrder, CThreadTransferSrcDstVectorDim, CThreadTransferDstScalarPerVector >

Namespaces
namespace	ck

Functions
template<typename GridwiseGemm, typename FloatAB, typename FloatC, typename ABK0MK1GridDesc, typename BBK0NK1GridDesc, typename CM0N0M1N1M2M3M4N2GridDesc, typename AElementwiseOperation, typename BElementwiseOperation, typename CElementwiseOperation, typename CBlockClusterAdaptor, bool HasMainKBlockLoop>
__global__ void	ck::kernel_gemm_xdlops_v2r4 (const FloatAB *__restrict__ p_a_grid, const FloatAB *__restrict__ p_b_grid, FloatC *__restrict__ p_c_grid, const ABK0MK1GridDesc a_b_k0_m_k1_grid_desc, const BBK0NK1GridDesc b_b_k0_n_k1_grid_desc, const CM0N0M1N1M2M3M4N2GridDesc c_m0_n0_m1_n1_m2_m3_m4_n2_grid_desc, const AElementwiseOperation a_element_op, const BElementwiseOperation b_element_op, const CElementwiseOperation c_element_op, const CBlockClusterAdaptor c_block_cluster_adaptor)