namd2.7b1 slower with openmpi 1.3.2 than namd2.6 with openmpi 1.2.6

From: Vlad Cojocaru (Vlad.Cojocaru_at_eml-r.villa-bosch.de)
Date: Thu Jul 02 2009 - 10:41:23 CDT

Dear namd users,

I compiled the cvs code of namd2.7b1 (checked out on 1st of July) with
openmpi 1.3.2 on a x86_64 architecture (quad cores opterons with
infiniband) and I noticed that it was 50% slower than my old compilation
of namd2.6 with openmpi 1.2.6 . New compilation was done with intel 11.0
compiler (-g -O2 flags), old compilation with intel 10.1.015 (same flags).

Then I compiled with mvapich2-1.4rc1 and I noticed that the 2.7b1
version it was about 10% to 15% faster than any of my old compilations
of namd2.6.

Both openmpi and mvapich2 I compiled myself with infiniband support
using same compiler. the flags used for the compilation of openmpi 1.3.2
were exactly the same as those used for the old compilation of openmpi
1.2.6.

Is there a reason for the drop in performance when compiled with openmpi
1.3.2 ?

Best wishes
Vlad

-- 
----------------------------------------------------------------------------
Dr. Vlad Cojocaru
EML Research gGmbH
Schloss-Wolfsbrunnenweg 33
69118 Heidelberg
Tel: ++49-6221-533202
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 - 15:51:09 CST