Shucai Xiao 2729ae6c6f use different int8 mfma instructions on different GPUs. (#368) ...

* changes support to choose different int8 instructions

* rename an instruction name

Co-authored-by: Aleksandr Efimov <efimov.alexander@gmail.com>

2023-10-25 19:12:21 -05:00

..

NVGPU

[BACKEND] Remove HopperHelpers.c and replace with inline ptx and LLVM codegen (#2047)

2023-08-10 15:52:37 -07:00

Triton

Initial commit to resolve merge conflicts

2023-10-03 04:04:26 +00:00

TritonGPU

use different int8 mfma instructions on different GPUs. (#368)

2023-10-25 19:12:21 -05:00

TritonNvidiaGPU

[BACKEND] Optimize performance for f16 epilogue with TMA store (#2135)

2023-08-21 12:44:11 -07:00

CMakeLists.txt

Initial code merge of Hopper support (#2036)

2023-08-07 09:53:04 +08:00