Advice for fast paraview rendering
 

Hi all,

at the moment I have a bit more time setting up my system and server for paraview and openfoam. Openfoam is still set up very good with optimal compiler flags for my hardware. Paraview is still compiled with the standard settings out of the thirdPartys.

I am interested in fast volume rendering and therefor I have two options:

• Using my Desktop-PC with graphic card and 8 core cpu
• Using my Server with onboard graphic card (very bad) and 24 cpu cores

Normally I run bigger cases on my server then copying the files to my desktop-pc and make my post-processing there. I think it would be more effective to make this with MESA-OS and my server instead of copying everything to my desk. pc, or?

I am not sure which one is faster but I think the server could be able to render faster than my desktop pc. Can someone give me some advice and links to inform me about the settings and prerequisites I have to do/have.
Maybe the rendering will speed up on my computer if i use off-screen rendering?



Summary: I want to have a fast rendering.

Thanks in advance,
Regards Tobi

Hi Tobi,

You've asked about this a few times in a few threads and this was the thread that I ended up finding. This week I saw this post on the VTK development mailing list: Overhaul of the rendering subsystem in VTK: Volume Rendering update - as you can deduce from this email, it's only natural that you're not getting the ultimate performance in volume rendering out of your hardware.

In case you don't know, the closest ParaView version to work properly with this reported volume rendering update should either be ParaView 4.1.0 or 4.2.0, because these two are the closest to using the latest VTK 6.x (6.1 release and 6.2 that is in development, if I'm not mistaken).
Therefore, if you have the patience to do it, you can try getting the VTK branch they mention on that email and use it to replace the VTK folder inside the ParaView source code folder and rebuild it from source code!

Best regards,
Bruno