2021.07.22 Meeting Notes

Agenda

Sparse interface PR
Ran into issue with DtoH and HtoD CUDA profiler output making CI fail
Need look into regression test suite now failing because of missing new gold (probably)
Added some text to paper

Will look further into communication patterns and load balancing
Looking at Parthenon code and comparing back to Athena++ paper
Question: All calculations on AMR octree are local, correct? Yes (tree construction and rank assignment are all done locally and identically on each rank)

Heads up: Prototyping a benchmarking tool suite that should include Parthenon at some point, also for downstream codes (sending out some slides)

Question: Are LANL downstream codes doing automated integration tests? No, only manual at this point

Participated in astrophysics conferences, not much time on Parthenon
Jonah and Ben have Monte Carlo transport running on Phoebus on top of Parthenon, pushing particles
Working on opacity library for use on GPUs (separate public library: https://github.com/lanl/singularity-opac)
Added text to paper. Are we adding code snippets?

Tracking down 0's (see Matrix channel), should be hopefully fixed now
Reviewed PRs
Working on adding AMR performance test to regression suite
Looks like we're getting new CI machine soon, current one is very dated. With new CI machine add Intel and Clang compilers
More interest in using AthenaPK for actual science applications and pleased with performance, results in more user support requirements and potential postdoc opportunities

Should we add code snippets and if so how to standardize it? Will depend on journal. SoftwareX? Super Computing (deadline soon)? Favorite right now seems to be The International Journal of High Performance Computing Applications (https://journals.sagepub.com/editorial-board/hpc). Decision: Let's go with The International Journal of High Performance Computing Applications (we're already using their Overleaf template)
Writing assignments still stand
Need to discuss scaling tests

Do release now (after fixing current regression tests)
Need some naming clean up and removal of old machinery pieces for next release