Difference between revisions of "Teleconference 2019-05-01"

From VTKM
Jump to navigation Jump to search
(Created page with "Attendees: ==ECP Updates== * Need to bias the KPP-3 Integration metric with a positive current value (what we have done so far) ** Hank working on this * Basis of Estimate d...")
 
 
Line 1: Line 1:
−
Attendees:
+
Attendees: Ken Moreland (SNL), Rob Maynard (Kitware), Matt Larsen (LLNL), Nick Davis (SNL), Hank Childs (UO), Abhishek Yenpure (UO)
  
 
==ECP Updates==
 
==ECP Updates==
  
 
* Need to bias the KPP-3 Integration metric with a positive current value (what we have done so far)
 
* Need to bias the KPP-3 Integration metric with a positive current value (what we have done so far)
−
** Hank working on this
+
** Should be good for now.
 
* Basis of Estimate due today!
 
* Basis of Estimate due today!
 
* Starting to plan work detail for FY20 and beyond
 
* Starting to plan work detail for FY20 and beyond
Line 36: Line 36:
 
* [MS-20/05] Cell Metrics
 
* [MS-20/05] Cell Metrics
 
** Brent and Daniel are working on an MR.
 
** Brent and Daniel are working on an MR.
 +
** Steven Walton is getting engaged on this.
 
** Initial MR will only be about 1/3 of the metrics.
 
** Initial MR will only be about 1/3 of the metrics.
 
** Will advance after that.
 
** Will advance after that.
Line 43: Line 44:
 
* [MS-20/07] Advanced Flow Algorithms
 
* [MS-20/07] Advanced Flow Algorithms
 
** Merged in fix to do double precision correctly.
 
** Merged in fix to do double precision correctly.
−
** Still deciding what we want to do.
+
** Going to spend this time getting the infrastructure rock solid.
−
*** Planning on talking during DOECGF
+
** Oregon will push on FTLE.
−
*** Hybrid parallel seems like a good way to go
+
** ORNL will work on stream surfaces and Poincare plot
−
** Organizing work with ALPINE and others
+
** Might not get all of them, but should at least succeed at one.
 +
** Also getting distributed memory parallelization working.
  
 
==Rountable notes==
 
==Rountable notes==
Line 57: Line 59:
 
* CEED Annual Meeting
 
* CEED Annual Meeting
 
** August 6-8
 
** August 6-8
 +
** Hank planning on going.
 +
** Probably Matt and Cyrus too
  
 
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.
 
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.
Line 65: Line 69:
  
 
Slides from DOECGF panel. (Highlights of VTK-m at scale we can send to Jim.)
 
Slides from DOECGF panel. (Highlights of VTK-m at scale we can send to Jim.)
 +
 +
Rob will take over meeting May 29 to talk to Thrust folks.
 +
 +
Sujin had been cleaning up memory leaks.
 +
 +
Removed DeviceAdapterError. It is no longer needed now that we have a default device adapter.
 +
 +
Sent some bugs to clang compiler team about issues compiler VTK-m.
 +
 +
Berk working on a library with the goal of loading VTK-m datasets from ADIOS. The idea is to create a library to define various types of schemas to various VTK-m data sets. Making progress. Can currently load an unstructured grid with a single cell type.
 +
 +
Matt has been working on generalized distributed memory interface.
 +
 +
Matt has been compiling things on Summit. Been struggling with the XLC compiler. Altered the VTK-m SPACK package to always set the NVCC native compiler.

Latest revision as of 14:06, 1 May 2019

Attendees: Ken Moreland (SNL), Rob Maynard (Kitware), Matt Larsen (LLNL), Nick Davis (SNL), Hank Childs (UO), Abhishek Yenpure (UO)

ECP Updates

ECP deliverable updates

FY19Q3

  • [MS-19/08] Point Movement 2
    • Done except user's guide.

FY19Q4

  • [MS-19/09] Path Geometry
    • Ramping up
    • Some of the functionality is already there
    • Probably need to add a poly-line cell type
  • [MS-20/01] Lightweight Cell Library
    • Been doing some changes on the interface to the VTK-m cell library.
    • Most of the underlying library is ready.
    • Almost ready.
    • Still need to implement VTK-m data around it.
  • [MS-20/02] Specialized Data Models
    • Dave P and Sujin talked about how to roll in some generalized extruded meshes.
  • [MS-20/03] Feature-Sensitive Surface Norm
    • Bumped up CUDA version so that we have our required CUDA 9
    • Can get back to actual algorithm
  • [MS-20/04] VTK-m Release 3
    • On track
    • Probably June would be a good time for a release.
  • [MS-20/05] Cell Metrics
    • Brent and Daniel are working on an MR.
    • Steven Walton is getting engaged on this.
    • Initial MR will only be about 1/3 of the metrics.
    • Will advance after that.
  • [MS-20/06] Contouring
    • Will model the problem after triangulation. Have different contouring for each cell type.
    • Ollie looking into VTK implementation and flying edge paper.
  • [MS-20/07] Advanced Flow Algorithms
    • Merged in fix to do double precision correctly.
    • Going to spend this time getting the infrastructure rock solid.
    • Oregon will push on FTLE.
    • ORNL will work on stream surfaces and Poincare plot
    • Might not get all of them, but should at least succeed at one.
    • Also getting distributed memory parallelization working.

Rountable notes

Upcoming events:

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

Hackathon?

Slides from DOECGF panel. (Highlights of VTK-m at scale we can send to Jim.)

Rob will take over meeting May 29 to talk to Thrust folks.

Sujin had been cleaning up memory leaks.

Removed DeviceAdapterError. It is no longer needed now that we have a default device adapter.

Sent some bugs to clang compiler team about issues compiler VTK-m.

Berk working on a library with the goal of loading VTK-m datasets from ADIOS. The idea is to create a library to define various types of schemas to various VTK-m data sets. Making progress. Can currently load an unstructured grid with a single cell type.

Matt has been working on generalized distributed memory interface.

Matt has been compiling things on Summit. Been struggling with the XLC compiler. Altered the VTK-m SPACK package to always set the NVCC native compiler.

Retrieved from "https://m-old.vtk.org/index.php?title=Teleconference_2019-05-01&oldid=703"