CUTLASS
CUDA Templates for Linear Algebra Subroutines and Solvers
Public Member Functions | Public Attributes | List of all members
cutlass::reduction::BatchedReductionTraits< ScalarA_, ScalarC_, ScalarD_, ScalarAlphaBeta_, ScalarAccum_, ReductionSize_, OutputTile_, SubTile_, ThreadShape_, Index_, BlockSwizzle_, maxInReg_, maxOutReg_, Functor_ >::Params Struct Reference

#include <batched_reduction_traits.h>

Collaboration diagram for cutlass::reduction::BatchedReductionTraits< ScalarA_, ScalarC_, ScalarD_, ScalarAlphaBeta_, ScalarAccum_, ReductionSize_, OutputTile_, SubTile_, ThreadShape_, Index_, BlockSwizzle_, maxInReg_, maxOutReg_, Functor_ >::Params:
Collaboration graph
[legend]

Public Member Functions

CUTLASS_HOST_DEVICE int initialize (Index m_, Index n_, ScalarAlphaBeta alpha_, ScalarAlphaBeta beta_, long long int reduction_stride_, ScalarA const *d_a_, Index lda_, ScalarC const *d_c_, Index ldc_, ScalarD *d_d_, Index ldd_)
 Initialize the parameters for 2D output tensor. More...
 

Public Attributes

Coord< 3 > problem_size
 The dimension of output tensor. More...
 
ScalarAlphaBeta alpha
 The alpha. More...
 
ScalarAlphaBeta beta
 The beta. More...
 
long long int reduction_stride
 stride between two element that will be sumed More...
 
ScalarA const * d_a
 
Index lda
 
ScalarC const * d_c
 
Index ldc
 
ScalarD * d_d
 
Index ldd
 
Functor::Params functorParams
 The functor params. More...
 

Member Function Documentation

template<typename ScalarA_ , typename ScalarC_ , typename ScalarD_ , typename ScalarAlphaBeta_ , typename ScalarAccum_ , int ReductionSize_ = 1, typename OutputTile_ = Shape<1, 1, 128>, typename SubTile_ = Shape<1, 1, 64>, typename ThreadShape_ = Shape<1, 1, 2>, typename Index_ = int, typename BlockSwizzle_ = DefaultBlockSwizzle, int maxInReg_ = 160, int maxOutReg_ = 64, typename Functor_ = typename cutlass::gemm::LinearScaling<ScalarAlphaBeta_, typename cutlass::gemm::FragmentMultiplyAdd<ScalarAlphaBeta_, ScalarAccum_, (ThreadShape_::kW % 2 == 0)> >>
CUTLASS_HOST_DEVICE int cutlass::reduction::BatchedReductionTraits< ScalarA_, ScalarC_, ScalarD_, ScalarAlphaBeta_, ScalarAccum_, ReductionSize_, OutputTile_, SubTile_, ThreadShape_, Index_, BlockSwizzle_, maxInReg_, maxOutReg_, Functor_ >::Params::initialize ( Index  m_,
Index  n_,
ScalarAlphaBeta  alpha_,
ScalarAlphaBeta  beta_,
long long int  reduction_stride_,
ScalarA const *  d_a_,
Index  lda_,
ScalarC const *  d_c_,
Index  ldc_,
ScalarD *  d_d_,
Index  ldd_ 
)
inline

Member Data Documentation

template<typename ScalarA_ , typename ScalarC_ , typename ScalarD_ , typename ScalarAlphaBeta_ , typename ScalarAccum_ , int ReductionSize_ = 1, typename OutputTile_ = Shape<1, 1, 128>, typename SubTile_ = Shape<1, 1, 64>, typename ThreadShape_ = Shape<1, 1, 2>, typename Index_ = int, typename BlockSwizzle_ = DefaultBlockSwizzle, int maxInReg_ = 160, int maxOutReg_ = 64, typename Functor_ = typename cutlass::gemm::LinearScaling<ScalarAlphaBeta_, typename cutlass::gemm::FragmentMultiplyAdd<ScalarAlphaBeta_, ScalarAccum_, (ThreadShape_::kW % 2 == 0)> >>
ScalarAlphaBeta cutlass::reduction::BatchedReductionTraits< ScalarA_, ScalarC_, ScalarD_, ScalarAlphaBeta_, ScalarAccum_, ReductionSize_, OutputTile_, SubTile_, ThreadShape_, Index_, BlockSwizzle_, maxInReg_, maxOutReg_, Functor_ >::Params::alpha
template<typename ScalarA_ , typename ScalarC_ , typename ScalarD_ , typename ScalarAlphaBeta_ , typename ScalarAccum_ , int ReductionSize_ = 1, typename OutputTile_ = Shape<1, 1, 128>, typename SubTile_ = Shape<1, 1, 64>, typename ThreadShape_ = Shape<1, 1, 2>, typename Index_ = int, typename BlockSwizzle_ = DefaultBlockSwizzle, int maxInReg_ = 160, int maxOutReg_ = 64, typename Functor_ = typename cutlass::gemm::LinearScaling<ScalarAlphaBeta_, typename cutlass::gemm::FragmentMultiplyAdd<ScalarAlphaBeta_, ScalarAccum_, (ThreadShape_::kW % 2 == 0)> >>
ScalarAlphaBeta cutlass::reduction::BatchedReductionTraits< ScalarA_, ScalarC_, ScalarD_, ScalarAlphaBeta_, ScalarAccum_, ReductionSize_, OutputTile_, SubTile_, ThreadShape_, Index_, BlockSwizzle_, maxInReg_, maxOutReg_, Functor_ >::Params::beta
template<typename ScalarA_ , typename ScalarC_ , typename ScalarD_ , typename ScalarAlphaBeta_ , typename ScalarAccum_ , int ReductionSize_ = 1, typename OutputTile_ = Shape<1, 1, 128>, typename SubTile_ = Shape<1, 1, 64>, typename ThreadShape_ = Shape<1, 1, 2>, typename Index_ = int, typename BlockSwizzle_ = DefaultBlockSwizzle, int maxInReg_ = 160, int maxOutReg_ = 64, typename Functor_ = typename cutlass::gemm::LinearScaling<ScalarAlphaBeta_, typename cutlass::gemm::FragmentMultiplyAdd<ScalarAlphaBeta_, ScalarAccum_, (ThreadShape_::kW % 2 == 0)> >>
ScalarA const* cutlass::reduction::BatchedReductionTraits< ScalarA_, ScalarC_, ScalarD_, ScalarAlphaBeta_, ScalarAccum_, ReductionSize_, OutputTile_, SubTile_, ThreadShape_, Index_, BlockSwizzle_, maxInReg_, maxOutReg_, Functor_ >::Params::d_a
template<typename ScalarA_ , typename ScalarC_ , typename ScalarD_ , typename ScalarAlphaBeta_ , typename ScalarAccum_ , int ReductionSize_ = 1, typename OutputTile_ = Shape<1, 1, 128>, typename SubTile_ = Shape<1, 1, 64>, typename ThreadShape_ = Shape<1, 1, 2>, typename Index_ = int, typename BlockSwizzle_ = DefaultBlockSwizzle, int maxInReg_ = 160, int maxOutReg_ = 64, typename Functor_ = typename cutlass::gemm::LinearScaling<ScalarAlphaBeta_, typename cutlass::gemm::FragmentMultiplyAdd<ScalarAlphaBeta_, ScalarAccum_, (ThreadShape_::kW % 2 == 0)> >>
ScalarC const* cutlass::reduction::BatchedReductionTraits< ScalarA_, ScalarC_, ScalarD_, ScalarAlphaBeta_, ScalarAccum_, ReductionSize_, OutputTile_, SubTile_, ThreadShape_, Index_, BlockSwizzle_, maxInReg_, maxOutReg_, Functor_ >::Params::d_c
template<typename ScalarA_ , typename ScalarC_ , typename ScalarD_ , typename ScalarAlphaBeta_ , typename ScalarAccum_ , int ReductionSize_ = 1, typename OutputTile_ = Shape<1, 1, 128>, typename SubTile_ = Shape<1, 1, 64>, typename ThreadShape_ = Shape<1, 1, 2>, typename Index_ = int, typename BlockSwizzle_ = DefaultBlockSwizzle, int maxInReg_ = 160, int maxOutReg_ = 64, typename Functor_ = typename cutlass::gemm::LinearScaling<ScalarAlphaBeta_, typename cutlass::gemm::FragmentMultiplyAdd<ScalarAlphaBeta_, ScalarAccum_, (ThreadShape_::kW % 2 == 0)> >>
ScalarD* cutlass::reduction::BatchedReductionTraits< ScalarA_, ScalarC_, ScalarD_, ScalarAlphaBeta_, ScalarAccum_, ReductionSize_, OutputTile_, SubTile_, ThreadShape_, Index_, BlockSwizzle_, maxInReg_, maxOutReg_, Functor_ >::Params::d_d
template<typename ScalarA_ , typename ScalarC_ , typename ScalarD_ , typename ScalarAlphaBeta_ , typename ScalarAccum_ , int ReductionSize_ = 1, typename OutputTile_ = Shape<1, 1, 128>, typename SubTile_ = Shape<1, 1, 64>, typename ThreadShape_ = Shape<1, 1, 2>, typename Index_ = int, typename BlockSwizzle_ = DefaultBlockSwizzle, int maxInReg_ = 160, int maxOutReg_ = 64, typename Functor_ = typename cutlass::gemm::LinearScaling<ScalarAlphaBeta_, typename cutlass::gemm::FragmentMultiplyAdd<ScalarAlphaBeta_, ScalarAccum_, (ThreadShape_::kW % 2 == 0)> >>
Functor::Params cutlass::reduction::BatchedReductionTraits< ScalarA_, ScalarC_, ScalarD_, ScalarAlphaBeta_, ScalarAccum_, ReductionSize_, OutputTile_, SubTile_, ThreadShape_, Index_, BlockSwizzle_, maxInReg_, maxOutReg_, Functor_ >::Params::functorParams
template<typename ScalarA_ , typename ScalarC_ , typename ScalarD_ , typename ScalarAlphaBeta_ , typename ScalarAccum_ , int ReductionSize_ = 1, typename OutputTile_ = Shape<1, 1, 128>, typename SubTile_ = Shape<1, 1, 64>, typename ThreadShape_ = Shape<1, 1, 2>, typename Index_ = int, typename BlockSwizzle_ = DefaultBlockSwizzle, int maxInReg_ = 160, int maxOutReg_ = 64, typename Functor_ = typename cutlass::gemm::LinearScaling<ScalarAlphaBeta_, typename cutlass::gemm::FragmentMultiplyAdd<ScalarAlphaBeta_, ScalarAccum_, (ThreadShape_::kW % 2 == 0)> >>
Index cutlass::reduction::BatchedReductionTraits< ScalarA_, ScalarC_, ScalarD_, ScalarAlphaBeta_, ScalarAccum_, ReductionSize_, OutputTile_, SubTile_, ThreadShape_, Index_, BlockSwizzle_, maxInReg_, maxOutReg_, Functor_ >::Params::lda
template<typename ScalarA_ , typename ScalarC_ , typename ScalarD_ , typename ScalarAlphaBeta_ , typename ScalarAccum_ , int ReductionSize_ = 1, typename OutputTile_ = Shape<1, 1, 128>, typename SubTile_ = Shape<1, 1, 64>, typename ThreadShape_ = Shape<1, 1, 2>, typename Index_ = int, typename BlockSwizzle_ = DefaultBlockSwizzle, int maxInReg_ = 160, int maxOutReg_ = 64, typename Functor_ = typename cutlass::gemm::LinearScaling<ScalarAlphaBeta_, typename cutlass::gemm::FragmentMultiplyAdd<ScalarAlphaBeta_, ScalarAccum_, (ThreadShape_::kW % 2 == 0)> >>
Index cutlass::reduction::BatchedReductionTraits< ScalarA_, ScalarC_, ScalarD_, ScalarAlphaBeta_, ScalarAccum_, ReductionSize_, OutputTile_, SubTile_, ThreadShape_, Index_, BlockSwizzle_, maxInReg_, maxOutReg_, Functor_ >::Params::ldc
template<typename ScalarA_ , typename ScalarC_ , typename ScalarD_ , typename ScalarAlphaBeta_ , typename ScalarAccum_ , int ReductionSize_ = 1, typename OutputTile_ = Shape<1, 1, 128>, typename SubTile_ = Shape<1, 1, 64>, typename ThreadShape_ = Shape<1, 1, 2>, typename Index_ = int, typename BlockSwizzle_ = DefaultBlockSwizzle, int maxInReg_ = 160, int maxOutReg_ = 64, typename Functor_ = typename cutlass::gemm::LinearScaling<ScalarAlphaBeta_, typename cutlass::gemm::FragmentMultiplyAdd<ScalarAlphaBeta_, ScalarAccum_, (ThreadShape_::kW % 2 == 0)> >>
Index cutlass::reduction::BatchedReductionTraits< ScalarA_, ScalarC_, ScalarD_, ScalarAlphaBeta_, ScalarAccum_, ReductionSize_, OutputTile_, SubTile_, ThreadShape_, Index_, BlockSwizzle_, maxInReg_, maxOutReg_, Functor_ >::Params::ldd
template<typename ScalarA_ , typename ScalarC_ , typename ScalarD_ , typename ScalarAlphaBeta_ , typename ScalarAccum_ , int ReductionSize_ = 1, typename OutputTile_ = Shape<1, 1, 128>, typename SubTile_ = Shape<1, 1, 64>, typename ThreadShape_ = Shape<1, 1, 2>, typename Index_ = int, typename BlockSwizzle_ = DefaultBlockSwizzle, int maxInReg_ = 160, int maxOutReg_ = 64, typename Functor_ = typename cutlass::gemm::LinearScaling<ScalarAlphaBeta_, typename cutlass::gemm::FragmentMultiplyAdd<ScalarAlphaBeta_, ScalarAccum_, (ThreadShape_::kW % 2 == 0)> >>
Coord<3> cutlass::reduction::BatchedReductionTraits< ScalarA_, ScalarC_, ScalarD_, ScalarAlphaBeta_, ScalarAccum_, ReductionSize_, OutputTile_, SubTile_, ThreadShape_, Index_, BlockSwizzle_, maxInReg_, maxOutReg_, Functor_ >::Params::problem_size
template<typename ScalarA_ , typename ScalarC_ , typename ScalarD_ , typename ScalarAlphaBeta_ , typename ScalarAccum_ , int ReductionSize_ = 1, typename OutputTile_ = Shape<1, 1, 128>, typename SubTile_ = Shape<1, 1, 64>, typename ThreadShape_ = Shape<1, 1, 2>, typename Index_ = int, typename BlockSwizzle_ = DefaultBlockSwizzle, int maxInReg_ = 160, int maxOutReg_ = 64, typename Functor_ = typename cutlass::gemm::LinearScaling<ScalarAlphaBeta_, typename cutlass::gemm::FragmentMultiplyAdd<ScalarAlphaBeta_, ScalarAccum_, (ThreadShape_::kW % 2 == 0)> >>
long long int cutlass::reduction::BatchedReductionTraits< ScalarA_, ScalarC_, ScalarD_, ScalarAlphaBeta_, ScalarAccum_, ReductionSize_, OutputTile_, SubTile_, ThreadShape_, Index_, BlockSwizzle_, maxInReg_, maxOutReg_, Functor_ >::Params::reduction_stride
The documentation for this struct was generated from the following file:
Generated by   doxygen 1.8.11