CUTLASS
CUDA Templates for Linear Algebra Subroutines and Solvers
Public Member Functions | Public Attributes | List of all members
cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params Struct Reference

Parameters structure.

#include <gemm_batched.h>

Collaboration diagram for cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params:
Collaboration graph
[legend]

Public Member Functions

CUTLASS_HOST_DEVICE Params ()
 
CUTLASS_HOST_DEVICE Params (cutlass::gemm::GemmCoord const &problem_size_, cutlass::gemm::GemmCoord const &grid_tiled_shape_, typename Mma::IteratorA::TensorRef ref_A_, int64_t stride_A_, typename Mma::IteratorB::TensorRef ref_B_, int64_t stride_B_, typename Epilogue::OutputTileIterator::TensorRef ref_C_, int64_t stride_C_, typename Epilogue::OutputTileIterator::TensorRef ref_D_, int64_t stride_D_, typename OutputOp::Params epilogue_, int batch_count_)
 

Public Attributes

cutlass::gemm::GemmCoord problem_size
 
cutlass::gemm::GemmCoord grid_tiled_shape
 
Mma::IteratorA::Params params_A
 
Mma::IteratorA::TensorRef ref_A
 
int64_t stride_A
 
Mma::IteratorB::Params params_B
 
Mma::IteratorB::TensorRef ref_B
 
int64_t stride_B
 
Epilogue::OutputTileIterator::Params params_C
 
Epilogue::OutputTileIterator::TensorRef ref_C
 
int64_t stride_C
 
Epilogue::OutputTileIterator::Params params_D
 
Epilogue::OutputTileIterator::TensorRef ref_D
 
int64_t stride_D
 
OutputOp::Params epilogue
 
int batch_count
 
int gemm_k_iterations
 

Constructor & Destructor Documentation

template<typename Mma_ , typename Epilogue_ , typename ThreadblockSwizzle_ >
CUTLASS_HOST_DEVICE cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params::Params ( )
inline
template<typename Mma_ , typename Epilogue_ , typename ThreadblockSwizzle_ >
CUTLASS_HOST_DEVICE cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params::Params ( cutlass::gemm::GemmCoord const &  problem_size_,
cutlass::gemm::GemmCoord const &  grid_tiled_shape_,
typename Mma::IteratorA::TensorRef  ref_A_,
int64_t  stride_A_,
typename Mma::IteratorB::TensorRef  ref_B_,
int64_t  stride_B_,
typename Epilogue::OutputTileIterator::TensorRef  ref_C_,
int64_t  stride_C_,
typename Epilogue::OutputTileIterator::TensorRef  ref_D_,
int64_t  stride_D_,
typename OutputOp::Params  epilogue_,
int  batch_count_ 
)
inline

Member Data Documentation

template<typename Mma_ , typename Epilogue_ , typename ThreadblockSwizzle_ >
int cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params::batch_count
template<typename Mma_ , typename Epilogue_ , typename ThreadblockSwizzle_ >
OutputOp::Params cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params::epilogue
template<typename Mma_ , typename Epilogue_ , typename ThreadblockSwizzle_ >
int cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params::gemm_k_iterations
template<typename Mma_ , typename Epilogue_ , typename ThreadblockSwizzle_ >
cutlass::gemm::GemmCoord cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params::grid_tiled_shape
template<typename Mma_ , typename Epilogue_ , typename ThreadblockSwizzle_ >
Mma::IteratorA::Params cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params::params_A
template<typename Mma_ , typename Epilogue_ , typename ThreadblockSwizzle_ >
Mma::IteratorB::Params cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params::params_B
template<typename Mma_ , typename Epilogue_ , typename ThreadblockSwizzle_ >
Epilogue::OutputTileIterator::Params cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params::params_C
template<typename Mma_ , typename Epilogue_ , typename ThreadblockSwizzle_ >
Epilogue::OutputTileIterator::Params cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params::params_D
template<typename Mma_ , typename Epilogue_ , typename ThreadblockSwizzle_ >
cutlass::gemm::GemmCoord cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params::problem_size
template<typename Mma_ , typename Epilogue_ , typename ThreadblockSwizzle_ >
Mma::IteratorA::TensorRef cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params::ref_A
template<typename Mma_ , typename Epilogue_ , typename ThreadblockSwizzle_ >
Mma::IteratorB::TensorRef cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params::ref_B
template<typename Mma_ , typename Epilogue_ , typename ThreadblockSwizzle_ >
Epilogue::OutputTileIterator::TensorRef cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params::ref_C
template<typename Mma_ , typename Epilogue_ , typename ThreadblockSwizzle_ >
Epilogue::OutputTileIterator::TensorRef cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params::ref_D
template<typename Mma_ , typename Epilogue_ , typename ThreadblockSwizzle_ >
int64_t cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params::stride_A
template<typename Mma_ , typename Epilogue_ , typename ThreadblockSwizzle_ >
int64_t cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params::stride_B
template<typename Mma_ , typename Epilogue_ , typename ThreadblockSwizzle_ >
int64_t cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params::stride_C
template<typename Mma_ , typename Epilogue_ , typename ThreadblockSwizzle_ >
int64_t cutlass::gemm::kernel::GemmBatched< Mma_, Epilogue_, ThreadblockSwizzle_ >::Params::stride_D
The documentation for this struct was generated from the following file:
Generated by   doxygen 1.8.11