mx roofline: adjust mxfp8 formulas #1953
Merged
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
Contributor
Author
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/1953
Note: Links to docs will display an error until the docs builds have been completed. This comment was automatically generated by Dr. CI and updates every 15 minutes. |
This was referenced Mar 25, 2025
facebook-github-bot
added
the
CLA Signed
vkuzo
added a commit
that referenced
this pull request
Mar 25, 2025
Summary: It's not clear whether we can write a fast dim0 + dim1 cast kernel, so adjusting the roofline estimation formulas to use separate dim0 and dim1 kernels Test Plan: ``` python benchmarks/float8/float8_roofline.py ~/local/tmp/20250325_b200_mxfp8_v2_triton.csv --mx_recipe_name mxfp8_cublas --shape_gen_name pow2_extended ``` Reviewers: Subscribers: Tasks: Tags: ghstack-source-id: 66a95b3 ghstack-comment-id: 2752441017 Pull Request resolved: #1953
vkuzo
added
the
topic: not user facing
This was referenced Mar 26, 2025
This was referenced Mar 26, 2025
jerryzh168
approved these changes
Mar 26, 2025
liangel-02
pushed a commit
that referenced
this pull request
Aug 25, 2025
* Update [ghstack-poisoned] * Update [ghstack-poisoned] * Update [ghstack-poisoned] * Update [ghstack-poisoned] * Update [ghstack-poisoned] * Update [ghstack-poisoned] * Update [ghstack-poisoned] * Update [ghstack-poisoned] * Update [ghstack-poisoned] * Update [ghstack-poisoned] * Update [ghstack-poisoned] * Update [ghstack-poisoned] * Update [ghstack-poisoned] * Update [ghstack-poisoned] * Update [ghstack-poisoned] * Update [ghstack-poisoned]
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.
Summary:
It's not clear whether we can write a fast dim0 + dim1 cast kernel, so
adjusting the roofline estimation formulas to use separate dim0 and dim1 kernels
Test Plan:
Reviewers:
Subscribers:
Tasks:
Tags: