[WIP] [AMD] Emit AMD specific intrinsics for dot #4594
Draft
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
This was referenced Aug 28, 2024
Closed
binarman
changed the title
|
This PR depends on #4516 |
alefimov-amd
force-pushed
the
v_dot_codegen
branch
from
b3f384c to
90a467a
Compare
|
Closing this PR for now. |
This PR introduces FMA dot operand converter and related tests.
- Fix compiler crashes in FMA.cpp - Fix lit test
- cleanup hash function in FMA.cpp - add more details in TODO in SharedToDotOperandFMA.cpp - cleanup DotOperandEncodingAttr::toLinearLayout
This PR: - Makes AccelerateAMDMatmul pass to emit FMA i8xi8->i32 and fp16xfp16->fp32 cases - Extends AMD FMA Dot code generation with new v_dot instructions for fp16xfp16 and int8 dtypes
binarman
force-pushed
the
v_dot_codegen
branch
from
90a467a to
8e03e70
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Outdated suggestions cannot be applied.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.
This PR:
This PR is a part of PR series. Final goal is to improve efficiency of small dot operations and bypass as much shared memory accesses as possible.Rough list of PRs:
Layout optimization, so operand B is loaded in proper mfma layout and do not need to go through LDS [WIP] Optimize fma dot #4581Vectorization optimization of dot operands/results (in case llvm can not do this internally)Reduction operation hoisting out of the K loop (reduction operation is a byproduct of layout optimization step) Hoist reduction outside a loop #4559