Looking to get any user experience from people who use/have used Paraview with Converge CFD. I am in a small company that uses Converge and currently weighing up switching to Paraview for all post-processing. We adopt python rather than Matlab so I think Paraview will fit in very well with the workflow. We use a small 48core HPC to run the simulations and to date use local machines for all post-processing, but I have noticed the NVIDIA IndeX for Paraview that I understand helps with running on a local machine rather than using any cores on the cluster?
Any information/experience from the group would be greatly appreciated,
For using NVIDIA IndeX on clusters for commercial purposes, you will need to get a license from NVIDIA (more details here).
However, note that IndeX is not a requirement for using ParaView. You can use ParaVIew on clusters using MPI for distributed computing. For such use-cases, one typically builds ParaView (w/o Qt, if that’s easier) on the HPC cluster using the MPI implementation deployed on the system. A matching version of the ParaView client downloaded from paraview.org can then be used on your desktop machine to connect to the HPC cluster.
ParaView includes support for distributed volume rendering out of the box as well.
Thank you for the reply Utkarsh, I think for now I will be concentrating on conducting any post-processing on local PCs, simply due to having limited CPUs on cluster. However I will certainly look to utilize the cluster in the future.