Refactor core scheduling algorithm to flag
Closed this issue · 1 comments
peterstanton commented
What kind of issue is this?
- Bug report
- Feature request
- Question not answered in documentation
- Cleanup needed
What is affected by this?
Performance
How do we replicate the issue/how would it work?
Currently, peter-multi-thread-sim manually distributes threads of execution to core. Refactor to scale based on hardware_concurrency() calls to get the number of calls, and activate when I use a command line flag. Then I can pull to the main multi-thread-sim branch.
Expected behavior (i.e. solution or outline of what it would look like)
BrainGrid should use regular hyperthreading unless I set the flag at the commandline.
Other Comments
stiber commented
Superseded by multi-cluster sim.