Teleconference 2019-08-07

Attendees: Ken Moreland (SNL), Rob Maynard (Kitware), Abhishek Yenpure (UO/SNL), Sujin Philip (Kitware), Allie Vacanti (Kitware)

ECP Updates

• ECP Annual Meeting
  • Proposals for sessions/tutorials is sooner than you think
• ECP review coming up
  • September 23-27
  • 30 minute talk like previous years
  • Will ask for input as necessary
• Capability Assessment Report (CAR) also coming up
  • Will update from last year
• KPP-3 will undergo another iteration
  • There will be a standard scoring system
  • It will likely be similar, but not accumulative

ECP deliverable updates

FY19Q4

• [MS-19/09] Path Geometry
  • Merged in a tube worklet
  • Still need to wrap that in a filter
  • Would like to clean poly data
• [MS-20/01] Lightweight Cell Library
  • Ran into an issue where we are using an eigen library for some matrix libraries.
  • Is supposed to support CUDA, but for some functions it does not.
  • Looking into alternatives.
• [MS-20/02] Specialized Data Models
  • Mark Kim checked in extruded cell set.
  • That is probably sufficient for the ECP milestone. (Probably some further tweaks.)
  • Now need some documentation
    • Will happen after Mark Kim completes paper (couple weeks out)
• [MS-20/03] Feature-Sensitive Surface Norm
  • Merged into VTK-m.
  • Hit a snag with VTK.
    • Cannot reorder in a VTK-m
  • Need User's Guide documentation
• [MS-20/04] VTK-m Release 3
  • Released happened (v1.4)
  • User's Guide is out.
• [MS-20/05] Cell Metrics
  • Brent made a MR. Awaiting for approval.
    • Ken or Rob need to get on this.
  • More functionality will be add incrementally
• [MS-20/06] Contouring
  • Ollie has started coding. Has some design questions he will ask offline.
• [MS-20/07] Advanced Flow Algorithms
  • FTLE implementation is done
    • But it depends on a sequence of MRs before check in.
  • Abhishek is advancing on FTLE. Looking into differences between VisIt implementation.
  • Dave sketching out a stream surface algorithm
  • Distributed memory issues merged into VTK-h

Hackathon

The hackathon was held last week. Hackathon 2019-07-30

During a postmortem talk, we discussed having an Invoker wrapped into the Filter superclasses so that the syntax of running a worklet from a filter would be "this->Invoke(worklet, arg1, arg2,...);"

Next hackathon will probably be in Oregon in the spring.

Rountable notes

Hank at CEED meeting. "CEED has their act together." They have a whole group doing accelerator stuff. Their target is 80% of peak. Hank gave 30min talk on Alpine and VTK-m.

A potential issue in the future is setting limits on the amount of memory. Using Umpire might help.