ahmed-bsod/gemm a8w8 gluon #1684
Open
+984
−22
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
ahmed-bsod
force-pushed
the
ahmed-bsod/gemm_a8w8_gluon
branch
from
fa00faf to
ce1e6fa
Compare
Contributor
There was a problem hiding this comment.
Pull request overview
This PR adds gluon kernel implementations for GEMM_A8W8 operations, including both a standard version and a preshuffled weight version. The implementation supports int8 and FP8 (e4m3/e5m2) input types with various output types (bf16, fp16, fp32, int32).
- Added two new gluon kernel variants:
gemm_a8w8andgemm_a8w8_preshufflefor AMD CDNA4 architecture - Extended test coverage to validate all three implementations (triton, gluon, gluon_shuffle) across multiple data types and configurations
- Updated benchmark suite to support gluon implementations with command-line flags for easy performance comparison
Reviewed changes
Copilot reviewed 4 out of 4 changed files in this pull request and generated 9 comments.
| File | Description |
|---|---|
aiter/ops/triton/gluon/gemm_a8w8.py |
New file implementing gluon-based GEMM_A8W8 kernels with standard and preshuffled weight variants |
op_tests/triton_tests/gemm/basic/test_gemm_a8w8.py |
Extended test suite to parametrize over implementation types and added support for int32 output dtype |
op_tests/op_benchmarks/triton/bench_gemm_a8w8.py |
Added command-line arguments for gluon and shuffle flags to enable performance benchmarking |
aiter/ops/triton/configs/gemm/gluon/gfx950-GEMM-A8W8.json |
Configuration file with tuned kernel parameters for gfx950 architecture |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
ahmed-bsod
force-pushed
the
ahmed-bsod/gemm_a8w8_gluon
branch
2 times, most recently
from
af5de67 to
acc815a
Compare
ahmed-bsod
force-pushed
the
ahmed-bsod/gemm_a8w8_gluon
branch
6 times, most recently
from
d069611 to
52fcd9c
Compare
ahmed-bsod
force-pushed
the
ahmed-bsod/gemm_a8w8_gluon
branch
from
52fcd9c to
ea4a722
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Motivation
Added gluon kernels for GEMM_A8W8 and gluon preshuffled GEMM_A8W8
Test Plan
Ran the test_gemm_a8w8.py script to make sure all functional tests passing
Test Result
Tests passed 🔥