Difference between revisions of "Teleconference 2019-03-13"

From VTKM
Jump to navigation Jump to search
(Created page with "Attendees: ==ECP Updates== * P6 Activities due by the end of this week. Have a meeting on this immediately following. ==ECP deliverable updates== ====FY19Q2==== * [MS-19...")
 
 
Line 1: Line 1:
−
Attendees:  
+
Attendees: Ken Moreland (SNL), Ollie Lo (LANL), Abhishek Yenpure (UO), Matt Larsen (LLNL), Rob Maynard (Kitware), Dave Pugmire (ORNL), Hank Childs (UO), Dave Rogers (LANL), Nick Davis (SNL)
  
  
Line 10: Line 10:
 
====FY19Q2====
 
====FY19Q2====
 
* [MS-19/04] Ghost Cells
 
* [MS-19/04] Ghost Cells
−
** Merge request in for ghost cell removal how Matt needs them.
+
** Code is all in. Just need to update documentation.
−
** Some issues with the ghost array being UInt8.
+
** UInt8 has been added to the list of default types.
−
*** Probably should add UInt8 to common types, at least in the short term.
 
−
*** In the long term, we need a way to support a slow path for "unknown" types.
 
−
** Ghost cell filter for structured cells has been merged in.
 
 
* [MS-19/05] Merge Points
 
* [MS-19/05] Merge Points
−
** Ken has an implementation in a merge request.
+
** Implementation has been merged in.
−
** Working through CUDA compile error.
+
** Still need to update documentation.
 
* [MS-19/06] Connected Components
 
* [MS-19/06] Connected Components
−
** Expecting to have an MR either end of this week or beginning of next.
+
** Should have MR this week.
 
* [MS-19/07] Advect Time Varying
 
* [MS-19/07] Advect Time Varying
−
** Still having some issues with grid evaluators.
+
** Abhishek and Rob have been helping with issues with the grid evaluating code.
−
*** Rob will look into it.
+
** Fallback is to consider time varying code we already have merged in.
 +
*** Everything will work for uniform and structured.
 +
*** Abhishek will go ahead and write a unit test for the backup.
 +
** Proposal: work on grid evaluator stack issue one more week. If it does not work by next meeting, let's fall back.
  
 
====FY19Q3====
 
====FY19Q3====
Line 31: Line 31:
  
 
==Rountable notes==
 
==Rountable notes==
 +
 +
Abhishek and Dave P have been working through issues with the grid evaluator. Abhishek has submitted a paper to EGPGV.
 +
 +
Steven Walten is now officially on the project. He and Matt are working with SW4.
 +
 +
Brent Lessley defended his dissertation. Hank volunteers to create a highlight slide on some of the great work that Brent did: nontraditional algorithms.
 +
 +
Matt has been working with Ascent release. Had some compiling issues. Noted that you must use separable compilation. If you are using virtual functions on CUDA, they must all be in the same library.
 +
 +
Reminder: DOECGF 2019 is coming up!
 +
 +
Allie has been working on graph iteration for consistent normals. Also working on a CUDA 10.1 dashboard. There has been a strange regression in the math library on Windows. Still looking into that.
 +
 +
Sujin working on VTK/VTK-m data set integration for ALPINE.
 +
 +
Rob has been doing GTC prep. Will be presenting VTK-m next week. Fixed some issues that came up when making slides. Fixed some issues with Summit/XL compiler.

Latest revision as of 13:43, 13 March 2019

Attendees: Ken Moreland (SNL), Ollie Lo (LANL), Abhishek Yenpure (UO), Matt Larsen (LLNL), Rob Maynard (Kitware), Dave Pugmire (ORNL), Hank Childs (UO), Dave Rogers (LANL), Nick Davis (SNL)


ECP Updates

  • P6 Activities due by the end of this week. Have a meeting on this immediately following.

ECP deliverable updates

FY19Q2

  • [MS-19/04] Ghost Cells
    • Code is all in. Just need to update documentation.
    • UInt8 has been added to the list of default types.
  • [MS-19/05] Merge Points
    • Implementation has been merged in.
    • Still need to update documentation.
  • [MS-19/06] Connected Components
    • Should have MR this week.
  • [MS-19/07] Advect Time Varying
    • Abhishek and Rob have been helping with issues with the grid evaluating code.
    • Fallback is to consider time varying code we already have merged in.
      • Everything will work for uniform and structured.
      • Abhishek will go ahead and write a unit test for the backup.
    • Proposal: work on grid evaluator stack issue one more week. If it does not work by next meeting, let's fall back.

FY19Q3

Skipping

FY19Q4

Skipping

Rountable notes

Abhishek and Dave P have been working through issues with the grid evaluator. Abhishek has submitted a paper to EGPGV.

Steven Walten is now officially on the project. He and Matt are working with SW4.

Brent Lessley defended his dissertation. Hank volunteers to create a highlight slide on some of the great work that Brent did: nontraditional algorithms.

Matt has been working with Ascent release. Had some compiling issues. Noted that you must use separable compilation. If you are using virtual functions on CUDA, they must all be in the same library.

Reminder: DOECGF 2019 is coming up!

Allie has been working on graph iteration for consistent normals. Also working on a CUDA 10.1 dashboard. There has been a strange regression in the math library on Windows. Still looking into that.

Sujin working on VTK/VTK-m data set integration for ALPINE.

Rob has been doing GTC prep. Will be presenting VTK-m next week. Fixed some issues that came up when making slides. Fixed some issues with Summit/XL compiler.

Retrieved from "https://m-old.vtk.org/index.php?title=Teleconference_2019-03-13&oldid=676"