Teleconference 2017-05-24

Attendees: Ken Moreland (SNL), Manish Mathai (UO), Dave Pugmire (ORNL), Tom Otahal (SNL), Rob Maynard (Kitware), Sujin Philip (Kitware), James Kress (ORNL), David Camp (LBL), Tom Fogal (NVIDIA), Ollie Lo (LANL)

Manish looking into nested parallelism on CUDA.

Dave P. is working on testing streaming for an LDAV paper.

Sujin has been making changes to extract VOI worklet. It was missing some cases and more complicated than necessary.

Berk submitted merge request that enables VTK-m in SENSEI.

Rob is working on a task design for loop tiling. Should build a little faster and create smaller binary size. Virtual only for TBB. CUDA still under development.

Started a code formatting branch. It reads through the code base to make the format consistent.

New README and contributing guide. README is expanded to have an example code and links. Contributing guide covers the entire process of creating a repository, making a branch, etc., etc.

Marching Cubes normals happens in a pass by itself. This allows us to keep for each output cell what input cells things come from. Normals now come from gradients. Gradients are only computed for cells that are eligible.

Pat Fasil was working on halo center finding. Should have a merge request soon.