Teleconference 2019-07-10

Attendees: Ken Moreland (SNL), Abhishek Yenpure (SNL/UO), Dave Pugmire (ORNL), Matt Larsen (LLNL), Robert Maynard (Kitware), Allie Vacanti (Kitware), Nick Davis (SNL), Jieyang (ORNL), Sujin Philip (Kitware), Berk Geveci (Kitware)

ECP Updates

• 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
  • Not much progress recently, but will start working on this full time.
• [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
• [MS-20/03] Feature-Sensitive Surface Norm
  • Working on cleaning up some CUDA issues
  • Need to do some benchmarks
• [MS-20/04] VTK-m Release 3
  • Released happened (v1.4)
  • Ken working on getting a User's Guide out.
• [MS-20/05] Cell Metrics
  • Brent still working on MR. Planning to work on it on Friday.
  • 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
  • Abhishek has an external FTLE filter
    • Needs a flowmap into VTK-m
  • Dave sketching out a stream surface algorithm

Hackathon

The hackathon will be held in Albuquerque starting July 30. Everyone planning on attending should register. Go to Hackathon 2019-07-30 for details.

Hank is planning to create some slides for a tutorial at the beginning of the hackathon.

Rountable notes

We should try to collect some ECP data so that as we do reporting we can quickly generate relevant images. It would be great if we could also share across ALPINE and other vis projects.

• Terry Turton has collected some stuff from LANL. Not sure if can be sent out.
  • She might have a list of distributed data in different places

There have been some issues with CUDA runtime errors in the cell locator hierarchy for changes unrelated. We discussed whether that might be a stack size issue. There is an example in kd-tree test to change the stack size. Rob suggested trying to revert building cell locators in vtkm_cont library.

There was a long discussion on the naming conventions for topology maps. We all agreed that it should change.

Ken has had some thought experiments about what if we scrapped trying to do virtual methods. He has written up an issue with the idea and encourages folks to take a look at it. (Ken needs to update the issue to talk about multiplexing within the technical plan.)

Jieyang is looking into using 16-bit floating point precision.