Ctrl+K
Checkout, Frequently Asked Questions!

In Situ @ KAUST#

To facilitate the use of In Situ Visualization at KAUST, we have developed an example repository (In Situ Vignettes) that explains and demonstrates how to run different in situ packes on Ibex and Shaheen. Please check out this repository for best practices and examples.

In Situ Vignettes Repo

In Situ Vignettes#

Work in progress

Why Use In Situ Visualization#

  • Faster simulations/More simulations

  • Increase simulation resolution (time, spatial)

  • Keep what you need

  • Reduced I/O helps other users too

  • Reduced storage keeps performance optimal

  • Faster simulations/More simulations

  • Increase simulation resolution (time, spatial)

  • Keep what you need

  • Reduced I/O helps other users too

  • Reduced storage keeps performance optimal

In Situ Options at KAUST#

Inshimtu#

KVL has developed Inshimtu

  • An in situ ‘shim’

  • Designed for existing, unmodified simulations

  • Works with files written by simulation

  • Uses Catalyst and ParaView vis-pipelines

  • Low barrier to entry

  • Try-out in situ without commitment of creating a true in situ integration

Check out an article written with Kitware on the use of Inshimtu Inshimtu Article

Check out the Inshimtu Repo Inshimtu Repo

ParaView Catalyst#

KVL has done a workshop on basic Catalyst use, see our YouTube Channel

Other options for Bespoke In Situ#

If you are looking for a true in situ integration there are multiple options available

  1. ADIOS2

    • A full data management solution.

    • Can aid in better I/O performance, in situ, data reduction

    • https://adios2.readthedocs.io/en/latest/

  2. ParaView Catalyst / VisIt Libsim

    • Directly integrate ParaView of VisIt functionality into your simulation code

    • https://catalyst-in-situ.readthedocs.io/en/latest/index.html

    • https://www.visitusers.org/index.php?title=VisIt-tutorial-in-situ

  3. Ascent

    • Ascent is a many-core capable flyweight in situ visualization and analysis infrastructure

    • https://ascent.readthedocs.io/en/latest/

  4. VTK-m

    • VTK-m is a toolkit of scientific visualization algorithms for emerging processor architectures (many-core extension of VTK)

    • It is used by Ascent, but can be used with other libraries to enable in transit visualization

    • https://m.vtk.org/

On this page