Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Optimizations for Fugaku: Pinning, Compile flags #755

Merged
merged 7 commits into from
May 31, 2024
Merged

Conversation

krenzland
Copy link
Contributor

Draft: Needs to be merged with master, new pinning routines need to be checked on other clusters.

This PR adds changes needed to run on Fugaku. This includes adding support for NUMA aware pinning in cases when CPU cores are deactivated using the hot-plugging system of Linux: https://www.kernel.org/doc/Documentation/ABI/testing/sysfs-devices-system-cpu

Copy link
Contributor

@Thomas-Ulrich Thomas-Ulrich left a comment

Choose a reason for hiding this comment

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

LGTM

@davschneller
Copy link
Contributor

As a note: I tried this one out on LUMI as well—but it didn't show any offline cores there.

However, lstopo shows the usual 7 CPUs (instead of 8). ... Maybe long-term it would be even of use to implement hwloc support?

@davschneller davschneller marked this pull request as ready for review May 14, 2024 11:24
@davschneller
Copy link
Contributor

Even though this branch should be kind of ready, it's probably better to wait at least until the training next week is over before merging it (the area it modifies may cause trouble with Mac OS systems—if there's some preprocessor guard missing somewhere).

@davschneller davschneller added this pull request to the merge queue May 31, 2024
Merged via the queue into master with commit 3cb4872 May 31, 2024
22 of 24 checks passed
@davschneller davschneller deleted the fugaku branch May 31, 2024 20:07
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.

None yet

3 participants