Ticket #11724 (inprogress)
Performance improvement for IntegrateEllipsoids
Reported by: | Vickie Lynch | Owned by: | Vickie Lynch |
---|---|---|---|
Priority: | major | Milestone: | Release 3.5 |
Component: | Framework | Keywords: | |
Cc: | Blocked By: | ||
Blocking: | Tester: |
Description
Looking at the performance tests, I’ve noticed that the change to run IntegrateEllipsoids in parallel has yielded an decrease in speed. See ​http://builds.mantidproject.org/view/Tests/job/performance_tests_master/Master_branch_performance_tests/ MDAlgorithmsTest.IntegrateEllipsoidsTestPerformance.test_execution_events Conceptually, doing this part of the processing in parallel should not have had this effect. For 3.5 do you think you could profile the performance tests on this algorithm to find out where the real bottle necks are?
Change History
comment:2 Changed 5 years ago by Vickie Lynch
Refs #11724 fix spelling
Changeset: 8870bbd527965590cadcc214724de7aa1e6a6d1b
Note: See
TracTickets for help on using
tickets.
Refs #11724 more changes for thread safety
Changeset: 8e5d4a8c0bbe2e58ad5c281bc847350e7ae90d86