CUTLASS
CUDA Templates for Linear Algebra Subroutines and Solvers
|
#include <utility>
#include "cutlass/cutlass.h"
#include "cutlass/relatively_equal.h"
#include "cutlass/util/distribution.h"
#include "tensor_foreach.h"
Go to the source code of this file.
Namespaces | |
cutlass | |
cutlass::reference | |
cutlass::reference::device | |
cutlass::reference::device::kernel | |
Functions | |
template<typename Element > | |
__global__ void | cutlass::reference::device::kernel::BlockCompareEqual (int *equal, Element const *ptr_A, Element const *ptr_B, size_t capacity) |
template<typename Element > | |
__global__ void | cutlass::reference::device::kernel::BlockCompareRelativelyEqual (int *equal, Element const *ptr_A, Element const *ptr_B, size_t capacity, Element epsilon, Element nonzero_floor) |
template<typename Element > | |
bool | cutlass::reference::device::BlockCompareEqual (Element const *ptr_A, Element const *ptr_B, size_t capacity, int grid_size=0, int block_size=0) |
Performs a bit-level equality check between two blocks. More... | |
template<typename Element > | |
bool | cutlass::reference::device::BlockCompareRelativelyEqual (Element const *ptr_A, Element const *ptr_B, size_t capacity, Element epsilon, Element nonzero_floor, int grid_size=0, int block_size=0) |
Performs a bit-level equality check between two blocks. More... | |