Alexander Efimov 6691de65db [MFMA] Support BFloat16 on MI100 (#295) ...

* [MFMA] Support BFloat16 on MI100

This PR makes use of mfma_f32_32x32x4bf16 instruction, available on MI100.

* fix tests, fix mfma encoding comment, fix switch between mfma versions.

* replace kDim from mfma layout with kWidth from dotOp layout

* rebase fix

* fix mfma to dot op shortcut for bfloat16

* fix review comments

2023-09-08 15:08:34 -05:00

..

examples

[FORMAT] autopep8 now uses max-line-length=88 (#1410)

2023-03-25 15:46:50 -07:00

src

[MLIR] Added tritongpu-stream-pipeline pass (#305)

2023-09-07 15:24:59 -05:00

test

[MFMA] Support BFloat16 on MI100 (#295)

2023-09-08 15:08:34 -05:00

triton

[MFMA] Support BFloat16 on MI100 (#295)

2023-09-08 15:08:34 -05:00

tutorials

refine gemm tuning scripts (#309)

2023-09-07 08:09:11 -05:00

MANIFEST.in

[TOOLS] improved ahead-of-time compiler (#1805)

2023-06-21 01:02:58 -07:00

pyproject.toml

[FRONTEND] Rename tl.reduction -> tl.reduce and improve testing (#1521)

2023-04-14 14:35:31 -07:00

setup.py

[BUILD] Reverted ptxas change and fixed bug in cache key computation (#1971)

2023-07-19 20:58:24 -07:00