Frame System
Andreas Zehender
az at azweb.de
Wed Oct 4 21:32:04 CEST 2006
Hi Miguel,
Miguel Angel Gómez Márquez schrieb:
> The problem for example when using POVRay in parallel the bottleneck
> is the reading of the file all cluster jobs depends on and this is a
> known issue presented (found) by the developer of MPI POVRay.
> This is why i give it so much importance.
> It is not a drawback at all in domestic use.
please write me some more details about this problem, I can't follow that.
From my experience with clusters in computing areas, network transfer
between different nodes was never a bottleneck so far. It never made
much differences if a mounted volume is located in the same machine or
in another machine. Or do you have any links and ressources how the
clustering in povray works?
Regards,
Andreas
More information about the kpovmodeler-devel
mailing list