Welcome, Guest
Username: Password: Remember me

TOPIC: Running simulation on a cluster

Running simulation on a cluster 12 years 11 months ago #3353

  • konsonaut
  • konsonaut's Avatar
  • OFFLINE
  • openTELEMAC Guru
  • Posts: 413
  • Thank you received: 144
Hello,

I run Telemac-2D on a cluster and I'm doing some benchmark tests with a 15000 element mesh. So according to first results, the simulation with cores > 16 is slower than with 12 cores.

Is there a certain rule how many cores to use for a given number of elements?

Best regards,

Clemens
The administrator has disabled public write access.

Re: Running simulation on a cluster 12 years 10 months ago #3359

  • jmhervouet
  • jmhervouet's Avatar
Hello,

This is not very surprising, as 15000 elements is a rather small mesh. However our tests with many more processors on very big meshes show that the speed up can be still interesting up to a few hundreds of points per sub-domain. You can try the Malpasset test case in 2D (26000 elements), which can run in about 4 s on a 12-core machine (about 10 s here at EDF on a 8-core Linux). It could be that clusters are not so optimised.

With best regards,

Jean-Michel Hervouet
The administrator has disabled public write access.

Re: Running simulation on a cluster 12 years 10 months ago #3447

  • konsonaut
  • konsonaut's Avatar
  • OFFLINE
  • openTELEMAC Guru
  • Posts: 413
  • Thank you received: 144
Hello Jean-Michel,

so what I understand it is a trial and error procedure: do some simulations for some time steps with different number of processors and see which number is the most suitable for the case.
As you suggested I tried the Malpasset case which takes 15 sec using 8 cores, 10 sec using 12 cores and 9 sec with 16 cores. So these simulations are not so fast as your simulations, nevertheless I'm happy :)

Best regards,

Clemens
The administrator has disabled public write access.
Moderators: borisb

The open TELEMAC-MASCARET template for Joomla!2.5, the HTML 4 version.