Teleconference 2019-08-28

From VTKM
Revision as of 17:48, 28 August 2019 by Kmorel (talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

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

ECP Updates

  • Budget plus ups
    • Remember to send BOEs to Ken by tomorrow
  • ECP review coming up
    • September 23-27
      • Will schedule this tomorrow
    • 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
    • Code is merged in
    • Just need User's guide documentation
    • Dave plans to have the documentation done this week
  • [MS-20/01] Lightweight Cell Library
    • Working through cell types
    • Need to set up proper repository and dashboards
  • [MS-20/02] Specialized Data Models
    • Just need documentation from Mark
    • Dave will ask Mark to get it done this week
    • Just added a patch for serialization of extruded cell sets
  • [MS-20/05] Cell Metrics
    • Implementations for 18 metrics.
      • Not all tested thoroughly
    • New student Sam Schwartz is getting up to speed
  • [MS-20/06] Contouring
    • All 3D cells have been covered
    • Still need 2D and 1D
    • It means that we have to change from CellSetSingleType to CellSetExplicit
      • Rob still wants to have a special case for structured cell sets
    • Added a tangle field source to help with testing
  • [MS-20/07] Advanced Flow Algorithms
    • Stream surface merged in a while ago
    • Dave plans to have the documentation done this week

Rountable notes

Allie has been working on ArrayGetValue(s). Also adding integer functionality to atomic arrays.

Rob working on fixing cuda compiler flags for downstream applications. Fixing some bugs in CMake itself.

Removed FilterCell because there is no need for it. Now just inherit from FilterField.

Berk had a discussion of moving distributed memory stuff from VTK-h to VTK-m. A stumbling block is the need to compile twice, once with MPI on and once with MPI off, but VTK-m takes too long to compile to compile it twice. Looking in to making a clean separation at the DIY layer so that only that small library needs to be compiled twice.

Setting up dashboards with new configurations (turning off 64-bit ids and turning on 64-bit float default).

Issue with multiple cell sets. The issue that has been brought up is that DatSet supports multiple CellSets. Each CellSet is identified by a name. Each cell field as to be associated by a particular CellSet by name. However, whenever a filter creates a new CellSet, it often changes the name (or more often ignores it). This makes all the CellSets invalid because they are no longer associated with a CellSet.

After a discussion, it was agreed that since we do not support multiple CellSets very well, we should remove that functionality. We can probably implement use cases using and upper level abstraction like multiblock.