Skip to content

Call 14 Jun 2017

Vivekanandan (Vivek) Balasubramanian edited this page Jun 14, 2017 · 3 revisions

Attendees:

Actions Review

Discussion Items:

  • Report about supported stages of the Princeton workflow
  • Report about ongoing effort to support Rhea
  • Report about ongoing effort to support GPU on Titan
  • Princeton SRS sign-off
  • Discuss PSU and Princeton science problems for the IPDPS paper
  • Discuss whether GPU support is within the scope of the IPDPS paper for Princeton science problem

Activity Reports:

AOB: None

Minutes:

  • PSU: discussing changes to their usecase, use case document is under edit mode. SRS is on hold.
  • PU usecase:
    • Initially useful to start with four scripts/components - forward simulations, analysis, adjoint simulations, post-analysis. It will be helpful to look into individual results. Integrating all of that into one script for automation can be one of the end goals.
    • Using the CPU version instead of the GPU version for the simulations might be possible but the experiments need to be downscaled. PU to internally discuss.
  • PU: to look at the SRS document to make sure everything is as expected so that we can sign-off
  • RU:
    • send initial EnTK script for forward simulations on Titan using CPU.
    • working towards Rhea support and GPU support
    • working on documentation
  • Discussion of experiments for the paper pending. To be picked up in the next meeting with more participants.

Clone this wiki locally