CUTLASS
CUDA Templates for Linear Algebra Subroutines and Solvers
|
Specialization for int8_t 4x4 transpose.
#include <transpose.h>
Public Types | |
using | TransposeShape = layout::PitchLinearShape< 4, 4 > |
using | Element = int8_t |
using | Fragment = cutlass::Array< Element, kElementCount > |
Public Member Functions | |
CUTLASS_DEVICE void | transform (Fragment &dst, Fragment &src) |
Static Public Attributes | |
static const int | kElementCount = ElementCount_ |
using cutlass::transform::thread::Transpose< ElementCount_, layout::PitchLinearShape< 4, 4 >, int8_t >::Element = int8_t |
using cutlass::transform::thread::Transpose< ElementCount_, layout::PitchLinearShape< 4, 4 >, int8_t >::Fragment = cutlass::Array<Element, kElementCount> |
using cutlass::transform::thread::Transpose< ElementCount_, layout::PitchLinearShape< 4, 4 >, int8_t >::TransposeShape = layout::PitchLinearShape<4,4> |
|
inline |
|
static |