[inductor] Skip addcmul decomposition to enable FMA lowering#175309
Closed
mlazos wants to merge 20 commits intogh/mlazos/106/basefrom
Closed
[inductor] Skip addcmul decomposition to enable FMA lowering#175309mlazos wants to merge 20 commits intogh/mlazos/106/basefrom
mlazos wants to merge 20 commits intogh/mlazos/106/basefrom
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/175309
Note: Links to docs will display an error until the docs builds have been completed. ❗ 1 Active SEVsThere are 1 currently active SEVs. If your PR is affected, please view them below: ✅ No FailuresAs of commit 1541f87 with merge base 0b6476f ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
This PR needs a
|
This was referenced Feb 19, 2026
This was referenced Feb 26, 2026
v0i0
approved these changes
Feb 27, 2026
Contributor
Author
|
@pytorchbot merge |
Collaborator
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
pytorchmergebot
pushed a commit
that referenced
this pull request
Feb 28, 2026
Add aten.addcdiv, aten.addcdiv_, and aten._foreach_addcdiv.Scalar to decomps_to_exclude so that the FMA-based addcdiv lowering is used instead of decomposition. Also simplify the dynamo handlers for addcdiv to always skip inline decomposition. This enables bitwise precision parity with eager CUDA for addcdiv operations used in Adam/AdamW optimizers. Pull Request resolved: #175310 Approved by: https://github.com/v0i0 ghstack dependencies: #174912, #175309
pytorchmergebot
pushed a commit
that referenced
this pull request
Mar 1, 2026
Eager CUDA computes `a + alpha * b` as `fma(b, alpha, a)`. Without this, Triton computes `b * alpha` then adds to `a` as separate operations, losing the FMA precision guarantee. This affects optimizer weight_decay paths which use `grad.add(param, alpha=weight_decay)` and `_foreach_add` with alpha. Authored with Claude. Pull Request resolved: #175838 Approved by: https://github.com/v0i0 ghstack dependencies: #174912, #175309, #175310
pytorchmergebot
pushed a commit
that referenced
this pull request
Mar 2, 2026
Eager CUDA computes `a + alpha * b` as `fma(b, alpha, a)`. Without this, Triton computes `b * alpha` then adds to `a` as separate operations, losing the FMA precision guarantee. This affects optimizer weight_decay paths which use `grad.add(param, alpha=weight_decay)` and `_foreach_add` with alpha. Authored with Claude. Pull Request resolved: #175838 Approved by: https://github.com/v0i0 ghstack dependencies: #174912, #175309, #175310
sandy-gags
pushed a commit
to sandy-gags/pytorch
that referenced
this pull request
Mar 12, 2026
Add aten.addcmul and aten._foreach_addcmul.Scalar to decomps_to_exclude so that the FMA-based addcmul lowering is used instead of decomposition. This enables bitwise precision parity with eager CUDA for addcmul operations. ghstack-source-id: 92d3504 Pull-Request: pytorch/pytorch#175309
EmanueleCoradin
pushed a commit
to EmanueleCoradin/pytorch
that referenced
this pull request
Mar 30, 2026
…#175309) Add aten.addcmul and aten._foreach_addcmul.Scalar to decomps_to_exclude so that the FMA-based addcmul lowering is used instead of decomposition. This enables bitwise precision parity with eager CUDA for addcmul operations. Pull Request resolved: pytorch#175309 Approved by: https://github.com/v0i0 ghstack dependencies: pytorch#174912
EmanueleCoradin
pushed a commit
to EmanueleCoradin/pytorch
that referenced
this pull request
Mar 30, 2026
…#175310) Add aten.addcdiv, aten.addcdiv_, and aten._foreach_addcdiv.Scalar to decomps_to_exclude so that the FMA-based addcdiv lowering is used instead of decomposition. Also simplify the dynamo handlers for addcdiv to always skip inline decomposition. This enables bitwise precision parity with eager CUDA for addcdiv operations used in Adam/AdamW optimizers. Pull Request resolved: pytorch#175310 Approved by: https://github.com/v0i0 ghstack dependencies: pytorch#174912, pytorch#175309
EmanueleCoradin
pushed a commit
to EmanueleCoradin/pytorch
that referenced
this pull request
Mar 30, 2026
Eager CUDA computes `a + alpha * b` as `fma(b, alpha, a)`. Without this, Triton computes `b * alpha` then adds to `a` as separate operations, losing the FMA precision guarantee. This affects optimizer weight_decay paths which use `grad.add(param, alpha=weight_decay)` and `_foreach_add` with alpha. Authored with Claude. Pull Request resolved: pytorch#175838 Approved by: https://github.com/v0i0 ghstack dependencies: pytorch#174912, pytorch#175309, pytorch#175310
EmanueleCoradin
pushed a commit
to EmanueleCoradin/pytorch
that referenced
this pull request
Mar 30, 2026
Eager CUDA computes `a + alpha * b` as `fma(b, alpha, a)`. Without this, Triton computes `b * alpha` then adds to `a` as separate operations, losing the FMA precision guarantee. This affects optimizer weight_decay paths which use `grad.add(param, alpha=weight_decay)` and `_foreach_add` with alpha. Authored with Claude. Pull Request resolved: pytorch#175838 Approved by: https://github.com/v0i0 ghstack dependencies: pytorch#174912, pytorch#175309, pytorch#175310
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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.
Stack from ghstack (oldest at bottom):
Add aten.addcmul and aten._foreach_addcmul.Scalar to decomps_to_exclude
so that the FMA-based addcmul lowering is used instead of decomposition.
This enables bitwise precision parity with eager CUDA for addcmul operations.
cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @ipiszy @kadeng @muchulee8 @amjames @chauhang @aakhundov @coconutruben @jataylo