[Paraview] Are there any standard benchmarks?

[Berk Geveci]

Hi folks,

I made a first pass at a benchmark script in ParaView. For now, it
creates a sphere and renders it using various settings - at 1M and 2M
polygons resolutions. It either prints the result or saves it as a
csv. To use it,

from the GUI, first connect to a server, then using the Python shell:

> from paraview import benchmark
> benchmark.run('results.csv')

from pvpython:

> from paraview import servermanager, benchmark
> servermanager.Connect() # change this to connect to a server if necessary
> benchmark.run('result.csv')

from pvbatch, either create a py file that has the above code or run
pvbatch as something like:

mpiexec -n 4 ./bin/pvbatch ./Utilities/VTKPythonWrapping/paraview/benchmark.py

This will print out the results to stdout.

Please do run the benchmark and send us the results. We can post them
on the wiki.

-berk

On Fri, Aug 1, 2008 at 3:45 AM, Ian Grimstead
<I.J.Grimstead at cs.cardiff.ac.uk> wrote:
> Hi All,
>
> I am highly suspicious of the rendering speed of our cluster (namely: a
> single low-end machine will beat an 8-node cluster, each note with an Nvidia
> Quadro card...).
>
> Are there any published benchmark results I can compare against? If not, it
> might be something to be considered, look at the Top 500 SuperComputers. Is
> there such a list for a rendering cluster? Bit of a distraction, but a good
> advert for ParaView :o)
>
> Cheers,
>
> Ian
> --------8<--------8<--------
> Dr. Ian Grimstead, Senior Research Associate, School of Computer Science,
> Cardiff University, 5 The Parade, Roath, Cardiff CF24 3AA, UK. Tel: +44 (0)
> 29 2087 9091
>
> _______________________________________________
> ParaView mailing list
> ParaView at paraview.org
> http://www.paraview.org/mailman/listinfo/paraview
>