[Paraview-developers] OSPRay high memory cost

[David E DeMarle]

Memory use is definitely higher than I would like it to be right now. Paul
Navratil and I have an IPCC going with Intel were we are looking at this
and related issues. Little progress so far yet I am afraid in identifying
what fat we can trim.

David E DeMarle
Kitware, Inc.
R&D Engineer
21 Corporate Drive
Clifton Park, NY 12065-8662
Phone: 518-881-4909 <(518)%20881-4909>

On Tue, Apr 11, 2017 at 8:19 AM, Favre Jean <jfavre at cscs.ch> wrote:

> Hello folks
>
> I am displaying particle datasets in the range of 0.3-3 billion particles
>
> Taking one dataset of 368,135,844 particles loaded as a PolyData with a
> single cell and a Poly-Vertex, with the default Point Rendering
> representation, the Unix "top" command shows that ParaView uses 13.8GB of
> memory. Expensive, but fair - I suppose. (No scalar attached, just
> coordinates).
>
> Switching on to OSPRay Point rendering, I get into massive swapping, and
> when the dust settles, I get interactive rendering, and ParaView uses 43.5
> GB of RAM. That's about 30GB for 368M particles.
>
> Is this a fair memory consumption?
>
> ospray-1.2.1.x86_64.linux
> embree-2.15.0.x86_64.linux
> ParaView master April 10, 2017
>
> -----------------
> Jean/CSCS
>
> _______________________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at http://www.kitware.com/opensou
> rce/opensource.html
>
> Search the list archives at: http://markmail.org/search/?q=
> Paraview-developers
>
> Follow this link to subscribe/unsubscribe:
> http://public.kitware.com/mailman/listinfo/paraview-developers
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/paraview-developers/attachments/20170413/a7ab9af4/attachment.html>