On 23/09/2011 4:37 PM, Sumanth M. Kumar wrote:
Hi everyone,
I did a simulation in an amazon cluster with 7.5 GB memory and 4 EC2
Compute Units (2 virtual cores with 2 EC2 Compute Units each) and it
was done with 0.5ns/day. Now Im running a simulation in another amazon
cluster of 7 GB of memory and 20 EC2 Compute Units (8 virtual cores
with 2.5 EC2 Compute Units each), but still the speed is same even if
i use "mpirun -np 8" with the mdrun, and finally its taking the same
time as 2core system, i.e. 0.5ns/day. I would like to know whether I
should make any changes during compilation? I had enabled mpirun
while compiling GROMACS. Do I need to make any changes, so that it
will consider all the 8 processors?
Virtual cores are useless for GROMACS because it saturates the processor
pipelines on its own (though Intel's hyperthreading from 5-6 years ago
was an exception). You need real cores, and to configure MPI to assign
processes 1-to-1 to real cores. The details will be up to you and your
system admins. You don't need to do anything with GROMACS other than
configure with MPI.
Mark
Thank you very much.
Sumanth
--
**
*Sumanth Kumar.M
*
On-the-Job Trainee
Plant Molecular Biology Lab.
Plant Breeding, Genetics and Biotechnology
International Rice Research Institute
DAPO-7777, Metro Manila
Philippines
Tel : +63- 9163129718
email :s.ku...@irri.org <mailto:s.ku...@irri.org>
sumanthkuma...@gmail.com <mailto:sumanthkuma...@gmail.com>
--
gmx-users mailing list gmx-users@gromacs.org
http://lists.gromacs.org/mailman/listinfo/gmx-users
Please search the archive at
http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-requ...@gromacs.org.
Can't post? Read http://www.gromacs.org/Support/Mailing_Lists