OpenMP scaling plateaus at ~8 threads #8

Open

Labels

UtkarshJoshiNtl

opened

the C++ propagator has a parallel-reduce bottleneck. Worth profiling and fixing for N > 10,000.

to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Assignees

No one assigned

Labels

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests