CK Tile Gemm API and heuristics changes #1809
Open
+1,130
−308
Conversation
This file contains 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
jakpiase
requested review from
junliume,
illsilin,
carlushuang,
qianfengz,
aosewski,
poyenc,
geyyer,
bartekxk,
andriy-ca and
afagaj
as code owners
jakpiase
changed the title
6 tasks
aosewski
requested changes
Jan 15, 2025
Comment on lines
+229
to
+236
| else if(a_layout == "C" && b_layout == "C") | ||
| { | ||
| return run_gemm_example_with_layouts(argc, argv, Col{}, Col{}, Row{}); | ||
| } | ||
| else if(a_layout == "C" && b_layout == "R") | ||
| { | ||
| return run_gemm_example_with_layouts(argc, argv, Col{}, Row{}, Row{}); | ||
| } |
There was a problem hiding this comment.
CCR and CRR don't work with the same tiles as RRR and RCR do, but with they seem to work with some other tile shapes, I am not sure if that's the desired behavior or if something is broken inside of them
aosewski
requested changes
Jan 20, 2025
| bool is_a_rowmajor; | ||
| bool is_b_rowmajor; | ||
| bool is_c_rowmajor; | ||
| std::string data_type; /** Tensors datatype, can be set to either fp16 or bf16*/ |
There was a problem hiding this comment.
fp8 should also be supported.
| @@ -2,6 +2,29 @@ | |||
| // Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. | |||
Comment on lines
+21
to
+39
| return gemm_<gemm_traits_<FP16, | ||
| FP16, | ||
| FP32, | ||
| FP16, | ||
| Row, | ||
| Row, | ||
| Row, | ||
| 256, | ||
| 256, | ||
| 32, | ||
| 2, | ||
| 2, | ||
| 1, | ||
| 32, | ||
| 32, | ||
| 16, | ||
| false, | ||
| false, | ||
| false>>(a, s); |
There was a problem hiding this comment.
Please turn off clang format also here. It would be beneficial to have // comment, as we used to have in old ck instances, describing what specific value referes to.
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.
Proposed changes
Add api for CK TILE Gemm similar to FMHA's one and add instances
Checklist
Please put an
xinto the boxes that apply. You can also fill these out after creating the PR. If you're not sure, please don't hesitate to ask.clang-formaton all changed filesDiscussion
Add api for CK Tile universal gemm and some instances for fp16 and bf16 cases. I'm not sure how the heuristic of choosing between comp and mem instances should be made so for now it is trivial and will be upgraded in the near future