Skip to content

Hypergraph#778

Open
lkotipal wants to merge 45 commits into
fmihpc:devfrom
lkotipal:hypergraph
Open

Hypergraph#778
lkotipal wants to merge 45 commits into
fmihpc:devfrom
lkotipal:hypergraph

Conversation

@lkotipal

@lkotipal lkotipal commented Jun 1, 2023

Copy link
Copy Markdown
Contributor
  • Add Zoltan Graph/Hypergraph support to Vlasiator, requires Hypergraph dccrg#27
  • Neighborhood used for hypergraph partitioning is parametrized as loadBalance.partitioning_neighborhood, defaulting to the full neighborhood encompassing extended sysboundaries and Vlasov translation.

@lkotipal

lkotipal commented Jun 1, 2023

Copy link
Copy Markdown
Contributor Author

image
So here are some profiling results. HYPERGRAPH seems to outperform RCB but loses to RIB in this test using samples/Magnetosphere3D on 8 Mahti nodes. Let me know if you want more numbers.

EDIT: numbers outdated, but results are still the same

@lkotipal lkotipal force-pushed the hypergraph branch 2 times, most recently from deaaeec to c97f20b Compare February 8, 2024 08:58
Comment thread grid.cpp Outdated

@markusbattarbee markusbattarbee left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

Some less and some more serious questions

Comment thread datareduction/datareducer.cpp
Comment thread ioread.cpp Outdated
Comment thread sysboundary/sysboundary.cpp Outdated
Comment thread vlasiator.cpp Outdated
Comment thread vlasovsolver/vlasovmover.cpp Outdated
@markusbattarbee

Copy link
Copy Markdown
Contributor

Neighborhood names need to be updated to be in accordance with the changes from #1020

@markusbattarbee markusbattarbee left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

some lingering depracted stuff from merges

Comment thread common.h Outdated
Comment thread datareduction/datareducer.cpp Outdated
Comment thread parameters.cpp Outdated
Comment thread ioread.cpp

@ursg ursg left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

This (and the corresponding DCCRG PR) looks all sane when reading throuh, and your results indicate that it at least doesn't break anything.

But I agree with Markus that a small at-scale test, especially in interaction with dAMR, should be done before merging.

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.

3 participants