CUTLASS
CUDA Templates for Linear Algebra Subroutines and Solvers
Public Member Functions | List of all members
cutlass::reference::host::Gemm< ElementA, LayoutA, ElementB, LayoutB, ElementC, LayoutC, ScalarType, ComputeType, arch::OpMultiplyAddSaturate > Struct Template Reference

Partial specialization for multiply-add-saturate.

#include <gemm.h>

Public Member Functions

void operator() (gemm::GemmCoord problem_size, ScalarType alpha, TensorRef< ElementA, LayoutA > tensor_a, TensorRef< ElementB, LayoutB > tensor_b, ScalarType beta, TensorRef< ElementC, LayoutC > tensor_c, ComputeType initial_accum=ComputeType(0))
 
void operator() (gemm::GemmCoord problem_size, ScalarType alpha, TensorRef< ElementA, LayoutA > tensor_a, TensorRef< ElementB, LayoutB > tensor_b, ScalarType beta, TensorRef< ElementC, LayoutC > tensor_c, TensorRef< ElementC, LayoutC > tensor_d, ComputeType initial_accum=ComputeType(0))
 

Member Function Documentation

template<typename ElementA , typename LayoutA , typename ElementB , typename LayoutB , typename ElementC , typename LayoutC , typename ScalarType , typename ComputeType >
void cutlass::reference::host::Gemm< ElementA, LayoutA, ElementB, LayoutB, ElementC, LayoutC, ScalarType, ComputeType, arch::OpMultiplyAddSaturate >::operator() ( gemm::GemmCoord  problem_size,
ScalarType  alpha,
TensorRef< ElementA, LayoutA >  tensor_a,
TensorRef< ElementB, LayoutB >  tensor_b,
ScalarType  beta,
TensorRef< ElementC, LayoutC >  tensor_c,
ComputeType  initial_accum = ComputeType(0) 
)
inline
template<typename ElementA , typename LayoutA , typename ElementB , typename LayoutB , typename ElementC , typename LayoutC , typename ScalarType , typename ComputeType >
void cutlass::reference::host::Gemm< ElementA, LayoutA, ElementB, LayoutB, ElementC, LayoutC, ScalarType, ComputeType, arch::OpMultiplyAddSaturate >::operator() ( gemm::GemmCoord  problem_size,
ScalarType  alpha,
TensorRef< ElementA, LayoutA >  tensor_a,
TensorRef< ElementB, LayoutB >  tensor_b,
ScalarType  beta,
TensorRef< ElementC, LayoutC >  tensor_c,
TensorRef< ElementC, LayoutC >  tensor_d,
ComputeType  initial_accum = ComputeType(0) 
)
inline

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