namd mpi version with smp

From: Vlad Cojocaru (Vlad.Cojocaru_at_eml-r.villa-bosch.de)
Date: Mon Aug 11 2008 - 05:22:18 CDT

Dear NAMD list,

I am testing a new cluster based on quad-cores AMD opterons. I was
thinking to test an mpi version of namd built with smp and compare the
speed with a pure mpi version ...

Does this make sense at all ? (I mean to use mpi to create 2 processes
on 2 different nodes and use the smp environment on 4 cores on each node
?) If this makes sense whatsoever, how would one run this? I was
thinking of "mpirun -n $nodes namd2 +p4". However this fails, as namd
appears to run only on 2 cores and also "+p4" after namd2 is taken as a
configuration file, thus namd crashes:

I am working with openmpi 1.2.6. I am not sure how is the smp support
for different mpi versions.

I built charm with mpi-linux-amd64-smp and namd on top of it ... Does
one need to add any threads options to the compiler while building namd
in this way ?

If somebody has experience with this, could you please comment a couple
of lines on it ?

Thank you very much

best wishes
vlad

-- 
----------------------------------------------------------------------------
Dr. Vlad Cojocaru
EML Research gGmbH
Schloss-Wolfsbrunnenweg 33
69118 Heidelberg
Tel: ++49-6221-533266
Fax: ++49-6221-533298
e-mail:Vlad.Cojocaru[at]eml-r.villa-bosch.de
http://projects.villa-bosch.de/mcm/people/cojocaru/
----------------------------------------------------------------------------
EML Research gGmbH
Amtgericht Mannheim / HRB 337446
Managing Partner: Dr. h.c. Klaus Tschira
Scientific and Managing Director: Prof. Dr.-Ing. Andreas Reuter
http://www.eml-r.org
----------------------------------------------------------------------------

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:49:44 CST