Teleconference 2023-02-22
Attendees: Ken Moreland (ORNL), Ollie Lo (LANL), Silvio Rizzi (ANL), Roxana Bujack (LANL), Manish Mathai (UO), Dave Pugmire (ORNL), Sujin Philip (Kitware), Jeff Amstutz (Nvidia), Abhishek Yenpure (Kitware), Hank Childs (UO), Berk Geveci (Kitware), Jay Wang (ORNL), Vicente Bolea (Kitware), Mark Bolstad (SNL), Tushar Athawale (ORNL)
ECP Updates
- Summit allocations
- There will no longer be automatic Summit allocations for ECP
- If we need Summit access for FY23/Q3, we must submit an allocation request by March 1
- Do we have any need/desire for Summit access? (Submission form: https://jira.exascaleproject.org/servicedesk/customer/portal/10/create/85)
- KPP-3
- Budget estimations for FY24
- Asking for 1/4 of FY23 spend plan for FY24Q1
- Was told by Jordan Williams that should not be an issue
- Trying to make sure contracts continue
- Asking for 1/4 of FY23 spend plan for FY24Q1
- Highlights
- Based on KPP-3 work
Porting Activities
- https://wiki.jlse.anl.gov/display/inteldga/ECP+2.3.4.13+VTK-m
- https://confluence.exascaleproject.org/display/STDA05/Porting+Activities+O21
ECP task updates
Roundtable
Bronson Messner approached Ken asking if VTK-m could do a KPP-3 run on Frontier (that did not need to stress the file system). After some discussion, it made sense to run the In Situ benchmark on a lot of processes. The benchmark needs to be fixed for MPI run. The two items to fix are a correct parallel rendering (image composite/reduction) and a correct partitioning of the input. Once that is done, it should be trivial to run the small program at scale.
Decided on doing 4 "early" KPP-3 metrics
- Flow for WarpX and/or WDMApp
- Springboard off Abhishek's thesis work
- Ascent integration
- ParaView integration
- VisIt integration
R&D100?
- Thinking of submitting for 2023
- Big stumbling block: needs to be "created" between Jan 2022 to March 2023
- Plan: Release VTK-m 2.0, do a copyright assertion
