CK Tile Gemm API and heuristics changes #1809
Open
+889
−307
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 20, 2025
aosewski
previously approved these changes
Jan 22, 2025
mozga-amd
reviewed
Jan 31, 2025
| using BDataType = ck_tile::half_t; | ||
| using AccDataType = float; | ||
| using CDataType = ck_tile::half_t; | ||
| // ToDo: Add more bias config to support different categories of GEMM. |
There was a problem hiding this comment.
Could you please add your nick name to todo?
| float gemm_(const ck_tile::GemmHostArgs& args, const ck_tile::stream_config& s); | ||
|
|
||
| /** | ||
| * \brief Invoke gemm function |
There was a problem hiding this comment.
Could you please use @ syntax instead of \. Please have a look at tile_partitioner file
| @@ -0,0 +1,124 @@ | |||
|
|
|||
| // SPDX-License-Identifier: MIT | |||
| // Copyright (c) 2024-2025, Advanced Micro Devices, Inc. All rights reserved. | |||
| @@ -101,6 +100,30 @@ float gemm_calc(const ck_tile::GemmHostArgs& args, const ck_tile::stream_config& | |||
| return ave_time; | |||
| } | |||
|
|
|||
| float gemm(const gemm_traits& t, const ck_tile::GemmHostArgs& args, const ck_tile::stream_config& s) | |||
| { | |||
| if(t.is_a_rowmajor && t.is_b_rowmajor && t.is_c_rowmajor) | |||
There was a problem hiding this comment.
[[Consider]] Please have a look that we repeatedly use mechanisms to check whether we have a specific layout... t.is_a_rowmajor && t.is_b_rowmajor && t.is_c_rowmajor etc.
Maybe it would be worth considering has_enabled_RRR mechanis *(1), either at the traits structure level or as a separate mechanism? This way, we could avoid duplicating these checks, reduce the possibility of mistakes etc.
*(1) Example:
t.has_enabled_RRR()
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