Skip to content

Conversation

@PatKamin
Copy link
Contributor

For better results stability, pin benchmark binaries to four cores with the maximum available frequency.

@PatKamin
Copy link
Contributor Author

Test run on the PVC perf machine: https://github.com/intel/llvm/actions/runs/18679758239/job/53258464661

@PatKamin PatKamin requested review from a team and vinser52 October 21, 2025 12:06
For better results stability, pin benchmark binaries to four cores with the maximum available frequency.
Copy link
Contributor

@lslusarczyk lslusarczyk left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

some suggestions

Copy link
Contributor

@lukaszstolarczuk lukaszstolarczuk left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@PatKamin
Copy link
Contributor Author

@intel/dpcpp-devops-reviewers, is adding numactl package to docker images ok?

CPU driver sometimes changes a bit the maximum frequency of selected cores. This, however, doesn't impact benchmark results greatly. Assumption of having first cores with highest frequency is enough.
@PatKamin
Copy link
Contributor Author

@intel/dpcpp-devops-reviewers, please review. This change will help stabilize benchmark results on the new BMG machine.

Note that benchmark job that actually uses numactl installed passed: https://github.com/intel/llvm/actions/runs/19104203407/job/55173686283.
The benchmark job that runs automatically for this PR doesn't have numactl package installed yet, so its result can be neglected.

As to the e2e test failure, it's unrelated.

@PatKamin
Copy link
Contributor Author

@intel/dpcpp-devops-reviewers, a friendly ping. This patch should also fix failing benchmarks on BMG machine: https://github.com/intel/llvm/actions/runs/19328176408/job/55285113356#step:19:2263

Copy link
Contributor

@lukaszstolarczuk lukaszstolarczuk left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

thx, LGTM

@sarnex
Copy link
Contributor

sarnex commented Nov 13, 2025

@PatKamin Is this ready for merge? Is the benchmark CI fail unrelated?

@lslusarczyk
Copy link
Contributor

@PatKamin Is this ready for merge? Is the benchmark CI fail unrelated?

Yes. As @PatKamin informed me, it is ready to merge and benchmarks fail will be fixed by dockers rebuilding this night.
Please merge.

@sarnex sarnex merged commit 20dd0bc into intel:sycl Nov 13, 2025
42 of 47 checks passed
@PatKamin PatKamin deleted the pin-4-cores-for-benches branch November 13, 2025 21:30
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

6 participants