2021.03.18 Meeting Notes

Agenda

• Individual/group updates
• Scaling update
• IO (performance and default basename handling)
• Argonne GPU Hackathon (pgrete)
• C++17 (pgrete)
• TODO
• Paper
• Next release
• Review non-WIP PRs

Individual/group updates

LANL CCS

Joshua S Brown

• Added Documentation for merging from forks
• Made some progress on generating performance metrics with Darshan for IO
• Opened up a pr to cleanup parameter inputs class
• Gitlab workshop
• Resolved issues with internal gitlab setup
• Still refactoring gitlab pr based on feedback

LANL Physics

Jonah Miller

Working on other projects

Ben Ryan

• Working through particle pr feedback, investigating possible mpi communication issues

AthenaPK

Philipp Grete

• Went through particle pr
• Added regression test sing leap from integrator
• Looking at particle accelerator, particles don't seem to be passing through meshblocks correctly.
• Has been looking into options for keeping kernels simple to avoid performance issues related over subscribing registers.

Forrest Glines

• Looked at static mesh refinement, looks like parthenon already has static mesh refinement implemented just needs to be documented.
• Looking into adding optional outputs to allow compatibility with Yt.

Jimm Stone

• Thinking through fix for solver's that Phil is working on.

Scaling updates

p9/volta seems to be crashing due to device memory allocations the p9 numbers are not really as great as we would like to see, Phil believes there is still room for an order of magnitude increase in performance. Problems, with the load balancing are causing the performance numbers to be unpredictable.

Scaling update

IO

IO performance metrics from RZAnsel

Argonne GPU Hackathon

C++17

TODO

• Finish profiling hdf5 io performance (Joshua Brown), use weak scaling, look at write rates.

Paper

Next Release