intrin_mfma_f32_4x4x4f16< 4, 64 > Struct Reference

intrin_mfma_f32_4x4x4f16&lt; 4, 64 &gt; Struct Reference#

Composable Kernel: ck::intrin_mfma_f32_4x4x4f16< 4, 64 > Struct Reference
ck::intrin_mfma_f32_4x4x4f16< 4, 64 > Struct Reference

#include <amd_xdlops.hpp>

Static Public Member Functions

template<class FloatC>
static __device__ void Run (const half4_t &reg_a, const half4_t &reg_b, FloatC &reg_c)

Member Function Documentation

◆ Run()

template<class FloatC>
__device__ void ck::intrin_mfma_f32_4x4x4f16< 4, 64 >::Run ( const half4_t & reg_a,
const half4_t & reg_b,
FloatC & reg_c )
inlinestatic

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