[rocSOLVER] Update rocsolver tests to measure median time#2455
Merged
Conversation
AGonzales-amd
added
the
noOptimizations
EdDAzevedo
reviewed
Nov 4, 2025
john00003
approved these changes
Nov 4, 2025
Contributor
john00003
left a comment
john00003
left a comment
There was a problem hiding this comment.
Looks good, thanks Angelo!
EdDAzevedo
approved these changes
Nov 4, 2025
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## develop #2455 +/- ##
============================================
+ Coverage 64.05% 75.25% +11.21%
============================================
Files 8 321 +313
Lines 3001 31198 +28197
Branches 445 4708 +4263
============================================
+ Hits 1922 23478 +21556
- Misses 861 4887 +4026
- Partials 218 2833 +2615
Flags with carried forward coverage won't be shown. Click here to find out more. 🚀 New features to boost your workflow:
|
tfalders
approved these changes
Nov 7, 2025
Contributor
tfalders
left a comment
tfalders
left a comment
There was a problem hiding this comment.
Looks good! Thanks @AGonzales-amd. Doesn't have to be this PR, but we should also look at using hip events to gather the timing data, and adding an argument to rocsolver-bench to select which averaging method to use.
EdDAzevedo
approved these changes
Nov 8, 2025
EdDAzevedo
approved these changes
Nov 12, 2025
assistant-librarian Bot
pushed a commit
to ROCm/rocSOLVER
that referenced
this pull request
Nov 13, 2025
[rocSOLVER] Update rocsolver tests to measure median time (#2455) ## Motivation The mean of gpu runtimes is too sensitive to outliers. This makes the results of `rocsolver-bench` inconsistent between runs on sensitive hardware. This PR changes the tests to compute the median instead of the mean. ## Technical Details Added `rocsolver_timer` class to control the behavior of collecting and combining execution times in the tests. ## Test Plan Compare before and after benchmark results (sytrd). ## Test Result No significant difference between mean and median. Therefore, the new time captured is representative of what we had prior. ## Submission Checklist - [x] Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.
ammallya
pushed a commit
that referenced
this pull request
Feb 3, 2026
* change cshuffle size * added mxfp4 moe async buffer loading without B preshuffle * added mx moe B shuffling + scale shuffling (async loads) * minor fix --------- Co-authored-by: mtgu0705 <mtgu@amd.com> [ROCm/composable_kernel commit: 7998ae8]
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
5 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.
Motivation
The mean of gpu runtimes is too sensitive to outliers. This makes the results of
rocsolver-benchinconsistent between runs on sensitive hardware. This PR changes the tests to compute the median instead of the mean.Technical Details
Added
rocsolver_timerclass to control the behavior of collecting and combining execution times in the tests.Test Plan
Compare before and after benchmark results (sytrd).
Test Result
No significant difference between mean and median. Therefore, the new time captured is representative of what we had prior.
Submission Checklist