Gate division bitwise numerics under a flag #165566
Closed
+32
−4
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
[ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/165566
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 3821222 with merge base f58f301 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy chenyang78 kadeng muchulee8 amjames chauhang aakhundov coconutruben [ghstack-poisoned]
#164144 ensures that division for compile is bitwise equivalent with eager. However, in #164301, the kernel performance is regressed. On B200: With standard triton `/`: 6511 GB/s With triton `div_rn`: 4692 GB/s Further investigation is required for the generated PTX to see why there is such a large slowdown. For now, enable bitwise equivalent results under `TORCHINDUCTOR_EMULATE_DIVISION_ROUNDING` similar to emulate_precision_cast cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy chenyang78 kadeng muchulee8 amjames chauhang aakhundov coconutruben [ghstack-poisoned]
ngimel
approved these changes
Oct 15, 2025
eellison
approved these changes
Oct 15, 2025
|
@pytorchbot merge |
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 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 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Outdated suggestions cannot be applied.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):
#164144 ensures that division for compile is bitwise equivalent with eager. However, in #164301, the kernel performance is regressed.
On B200:
With standard triton
/:6511 GB/s
With triton
div_rn:4692 GB/s
Further investigation is required for the generated PTX to see why there is such a large slowdown. For now, enable bitwise equivalent results under
TORCHINDUCTOR_EMULATE_DIVISION_ROUNDINGsimilar to emulate_precision_castcc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @ipiszy @chenyang78 @kadeng @muchulee8 @amjames @chauhang @aakhundov @coconutruben