[Paraview-developers] [EXTERNAL] Benchmarking Paraview

[Ken Martin]

Here is an email from Chuck from a while back, it might be useful

You should be able to run the many spheres benchmark script that ships with
ParaView.  You can run it with pvbatch:

[chuck at hal9000]$ pvbatch ${ParaView_DIR}/lib/paraview-5
.1/site-packages/paraview/benchmark/manyspheres.py --help
usage: manyspheres.py [-h] [-o OUTPUT_BASENAME] [-s SPHERES]
                      [-n SPHERES_IN_SCENE] [-r RESOLUTION] [-v VIEW_SIZE]
                      [-f FRAMES] [-c] [-y]

Benchmark ParaView geometry rendering

optional arguments:
  -h, --help            show this help message and exit
  -o OUTPUT_BASENAME, --output-basename OUTPUT_BASENAME
                        Basename to use for generated output files
  -s SPHERES, --spheres SPHERES
                        The total number of spheres to render
  -n SPHERES_IN_SCENE, --spheres-in-scene SPHERES_IN_SCENE
                        The number of spheres in the entire scene, including
                        those not rendered.
  -r RESOLUTION, --resolution RESOLUTION
                        Theta and Phi resolution to use for the spheres
  -v VIEW_SIZE, --view-size VIEW_SIZE
                        View size used to render
  -f FRAMES, --frames FRAMES
                        Number of frames
  -c, --color           Enable color renderings
  -y, --OSPRay          Use OSPRAY to render


The benchmark will generate several spheres in the shape of a cube and
sp[in the camera around them.  When running in parallel, each rank will be
processing a subset of the spheres.  A good value to use for "resolution"
is 726 since that gives you ~ 1M triangles per sphere.   When benching the
performance of a system, I generally find it easiest to keep the resolution
filed while varying the number of spheres.

For example, say you wanted to do a test run of 2048 spheres across 128
ranks (each rank renders 16 spheres) with ~ 1M triangles per sphere in a
1920x1080 view for 100 frames.  You would use:

mpirun -n 128 pvbatch ${ParaView_DIR}/lib/paraview-5
.1/site-packages/paraview/benchmark/manyspheres.py -s 2048 -r 726 -v
1920,1080 -f 100

This will output per-frame statistics across ranks as well as a final
summary of: Frames / Sec, Points / Frame, Polys / Frame, and MiPolys / Sec.

On Tue, Nov 29, 2016 at 3:32 PM, Scott, W Alan <wascott at sandia.gov> wrote:

> Only benchmarking tool I can think of is the Timer Log within ParaView.
> Tools/ Timer Log.
>
>
>
> Alan
>
>
>
> *From:* Paraview-developers [mailto:paraview-developers-
> bounces at paraview.org] *On Behalf Of *Khayam Gondal
> *Sent:* Tuesday, November 29, 2016 12:00 PM
> *To:* ParaView Developers List <paraview-developers at paraview.org>
> *Subject:* [EXTERNAL] [Paraview-developers] Benchmarking Paraview
>
>
>
> I have to benchmark paraview performance with different configurations
> such as with single vs multi GPU, with or without MPI support. I want to
> get results like GPU render time. Parallel IO timings. Server-Client
> communication latency.
>
>
>
> But i am kind of stuck here. I don't know if paraview have some
> benchmarking tools. Or whats is the best way to do such benchmarks?
>
> _______________________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at http://www.kitware.com/
> opensource/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
>
>


-- 
Ken Martin PhD
Chairman & CFO
Kitware Inc.
28 Corporate Drive
Clifton Park NY 12065
518 371 3971

This communication, including all attachments, contains confidential and
legally privileged information, and it is intended only for the use of the
addressee.  Access to this email by anyone else is unauthorized. If you are
not the intended recipient, any disclosure, copying, distribution or any
action taken in reliance on it is prohibited and may be unlawful. If you
received this communication in error please notify us immediately and
destroy the original message.  Thank you.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/paraview-developers/attachments/20161129/1318b5ad/attachment-0001.html>