[gmx-users] parallel computations
fabio.affinito at tiscali.it
Tue Jul 13 11:49:09 CEST 2004
I tried shorter runs, and I found out that parallel computations seems
to be slower than serial...
(Mnbf/s) (MFlops) (ps/NODE hour) (NODE hour/ns)
Performance: 8.449 670.583 5.217 191.678
Performance: 8.406 667.170 5.191 192.658
Performance: 1.882 154.785 1.162 860.278
Looking inside one of the logfiles of the last simulation I found:
Total NODE time on node 1: 1268.41
Average NODE time: 271.69
Load imbalance reduced performance to 200% of max
What does it means?
David van der Spoel wrote:
>On Tue, 2004-07-13 at 09:28, fabio affinito wrote:
>>I'm running a MPI simulation with Gromacs on 4 processors. The problem
>>is that only one of the four logfiles seems to be advancing (and not
>>very fast!), Is this normal?
>>This is the situation of my directory this morning at 9.40
>>-rw-r--r-- 1 affinito G09 2351 Jul 12 14:40 Simulazione3.log
>>-rw-r--r-- 1 affinito G09 2351 Jul 12 14:40 Simulazione2.log
>>-rw-r--r-- 1 affinito G09 2351 Jul 12 14:40 Simulazione1.log
>>-rw-r--r-- 1 affinito G09 45941 Jul 13 08:39 Simulazione0.log
>Yes this is normal, however you do have to check whether your
>calculation is efficient, i.e. do short test runs on 1,2,4 processors.
>>Thanks again.. and bye!
>>gmx-users mailing list
>>gmx-users at gromacs.org
>>Please don't post (un)subscribe requests to the list. Use the
>>www interface or send it to gmx-users-request at gromacs.org.
More information about the gromacs.org_gmx-users