diff options
| author | negargoli93 <[email protected]> | 2018-05-12 17:11:34 -0700 |
|---|---|---|
| committer | negargoli93 <[email protected]> | 2018-05-12 17:11:34 -0700 |
| commit | bae67e6a355047e360c30391588c2076913f86fa (patch) | |
| tree | 53ca7ce82640e29ca1ad1547379b28b87a6493ca /src/cuda-sim/instructions.cc | |
| parent | 68674d4ba230df0d3bf9f4e5b035f4cf9cfc185b (diff) | |
mma added
Diffstat (limited to 'src/cuda-sim/instructions.cc')
| -rw-r--r-- | src/cuda-sim/instructions.cc | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/src/cuda-sim/instructions.cc b/src/cuda-sim/instructions.cc index 7903343..bddaf8c 100644 --- a/src/cuda-sim/instructions.cc +++ b/src/cuda-sim/instructions.cc @@ -1565,7 +1565,7 @@ void mma_impl( const ptx_instruction *pI, core_t *core, warp_inst_t inst ) for (i=0;i<16;i++){ for(j=0;j<16;j++){ for(k=0;k<16;k++){ - matrix_d[i][j].f32=matrix_d[i][j].f32+matrix_a[i][k].f32*matrix_b[j][k].f32; + matrix_d[i][j].f32=matrix_d[i][j].f32+matrix_a[i][k].f32*matrix_b[k][j].f32; } matrix_d[i][j].f32+=matrix_c[i][j].f32; } |
