Skip to content

aggregation is a bottleneck #4

@krober10nd

Description

@krober10nd

...sending all data to 0 to write to disk is maybe not the best idea. Perhaps writing each processor's connectivity table to an HDF5 file in parallel may be more efficient?

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or request

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions