[gmx-users] VSMP cluster performance changes with time

Anupam Nath Jha anupam at mbu.iisc.ernet.in
Thu Aug 18 14:21:08 CEST 2011



I am using 24 processors and it is same for all steps.
this is the command in use:

----------------
tpbconv_mpi -f run.trr -s run.tpr -e run.edr -extend 1000 -o ext_11
nohup /opt/ScaleMP/mpich2/1.0.8/bin/mpirun -machinefile $PBS_NODEFILE -np 24
/opt/gromacs4/bin/mdrun_mpi -v -deffnm ext_11
--------------------------------------------------------------

similar for next steps.

regards
anupam



> On 18/08/11, Anupam Nath Jha <anupam at mbu.iisc.ernet.in> wrote:
>
>>
>>
>> Dear all
>>
>> I am running a MD (Gromacs - 4.5.3) in VSMP cluster with 24nodes.
>>
>> I am writing log files for different ns (1-10) in different files i.e.
>> different
>> files for each nano second . The performance goes like following:
>>
>> --------------------------
>> for 1st ns:
>>
>>                (Mnbf/s)   (GFlops)   (ns/day)  (hour/ns)
>> Performance:   1444.719     85.902      3.946      6.082
>> Finished mdrun on node 0 Tue Aug 16 21:46:02 2011
>>
>> at 2nd ns:
>>
>>                 (Mnbf/s)   (GFlops)   (ns/day)  (hour/ns)
>> Performance:    560.393     33.322      1.531     15.679
>> Finished mdrun on node 0 Wed Aug 17 13:27:38 2011
>> ------------------
>>
>> and in case of 3rd ns:
>>  it didn't complete the all steps and went in next.
>>
>> Is it the problem with the gromacs?
>>
>
>
> Probably not. Assuming you have kept everything else constant, it is likely that
> other activity on the machine or file system is delaying you somehow. However,
> in the absence of much basic information (like how many processors you are
> trying to use) we can't suppose much. Copying and pasting command lines is
> useful.
>
> Mark
>
> --
> This message has been scanned for viruses and
> dangerous content by MailScanner, and is
> believed to be clean.
>
> --
> gmx-users mailing list    gmx-users at 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-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/Support/Mailing_Lists


-- 
Science is facts; just as houses are made of stone, so is science is made of
facts; but a pile of stones is not a house, and  a collection of facts is not
necessarily science.

Anupam Nath Jha
Saraswathi Vishveshwara Lab
Molecular Biophysics Unit
IISc,Bangalore-560012
Karnataka
Ph. no.-22932611



-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.




More information about the gromacs.org_gmx-users mailing list