[gmx-users] Parallel run on a Quad Core

Justin A. Lemkul jalemkul at vt.edu
Sun Dec 21 17:45:09 CET 2008



Chitrita Dutta Roy wrote:
> tar -zxvf openmpi-1.2.8.tar.bz2
> cd openmpi-1.2.8
> ./configure
> make
> make install
> 
> i had my fftw library installed.
> 
> untarred gromacs went to its directory..
> ./configure --enable-mpi --program_suffix="_mpi"
> make
> make install
> 
> grompp -f pr2.mdp -c em.gro -r em.gro -p scr8.top -n em.ndx -o pr.tpr
> mpirun -np 4 /usr/local/gromacs/bin/mdrun -v -s pr.tpr -deffnm pr
> 

Well, that can't be right; all of your programs should have the _mpi suffix, if 
you followed the steps above.

What you've done isn't entirely necessary, although I don't know if it will have 
any adverse effects.  Follow the exact instructions posted here, they are given 
line-by-line and are perfectly clear:

http://www.gromacs.org/content/view/20/34/

> i tried like these besides i also tried with mdrun_mpi but the same 
> occured rather showed more time requirement to complete.And i would also 
> liked to know one thing if i want to do a fresh install removing the 
> earlier one.How would i remove the earlier one.As i tried doing it from 
> yum remove and rpm -e but all went in vain although they can show that 
> these things are installed.
> 

Just remove the Gromacs directory and try again, as long as any previous 
attempts haven't installed in other locations (i.e., /usr/local/bin, etc.)

-Justin

> Thanks again for your help.Plz instruct me to get one complete set of 
> steps to succefully install the Gromacs with mpi and to utilize the 
> power of four cores.
> 
> 
> On Sun, Dec 21, 2008 at 9:52 PM, Justin A. Lemkul <jalemkul at vt.edu 
> <mailto:jalemkul at vt.edu>> wrote:
> 
> 
> 
>     Chitrita Dutta Roy wrote:
> 
>         I tried that way it basically creates 4 jobs and takes hold of 4
>         cpus but the problem is that it shows to take more time than it
>         would normally without mpirun..and not only that a weird thing
>         starts happening.Once being progressed to some steps it retreats
>         back again to an older one and sometimes it just loops around in
>         that fashion calculating practically the same set of steps over
>         and over again.
> 
>         I somehow feel that the tpr file which is needed for the mdrun
>         was not built properly to run on 4 separate cores
>         simultaneously.ANd that is why maybe actually the 4 cores are
>         running the same copy of the program and so the same steps.And
>         what i see is kinda race condition.I guess i compiled it
>         properly with mpi so..now i am helpless.
> 
>         Thanks for your help.Waiting for some more.
> 
> 
>     Please post the following information:
> 
>     1. The exact series of commands you gave when installing Gromacs.
>     2. The exact commands you are issuing to create your job (grompp and
>     mdrun).
> 
>     -Justin
> 
> 
>         On Sun, Dec 21, 2008 at 7:56 PM, Manik Mayur
>         <manik.mayur at gmail.com <mailto:manik.mayur at gmail.com>
>         <mailto:manik.mayur at gmail.com <mailto:manik.mayur at gmail.com>>>
>         wrote:
> 
>            On Sun, Dec 21, 2008 at 5:04 PM, Chitrita Dutta Roy
>         <chitrita.md <http://chitrita.md>
>            <http://chitrita.md>@gmail.com <http://gmail.com>
>         <http://gmail.com>> wrote:
> 
>                Thanks to all for helping me previously.
>                Now i am unable to arrange the .tpr file for position
>         restraint
>                mdrun on a quad core as in version 4.0.2 grompp -np is not a
>                valid comand can anybody help me in building a tpr file for
>                parallel run on 4 cores of a quad core machine .I have
>         installed
>                gromacs using Open MPI Library and my files are suffixed
>         as _mpi.
> 
>                try running: mpirun -np 4 mdrun_mpi options
> 
> 
>                _______________________________________________
>                gmx-users mailing list    gmx-users at gromacs.org
>         <mailto:gmx-users at gromacs.org>
>                <mailto:gmx-users at gromacs.org <mailto:gmx-users at gromacs.org>>
> 
>                http://www.gromacs.org/mailman/listinfo/gmx-users
>                Please search the archive at http://www.gromacs.org/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
>         <mailto:gmx-users-request at gromacs.org>
>                <mailto:gmx-users-request at gromacs.org
>         <mailto:gmx-users-request at gromacs.org>>.
> 
>                Can't post? Read
>         http://www.gromacs.org/mailing_lists/users.php
> 
> 
> 
> 
>            --    _________________
>            HAPAX LEGOMENA
> 
>            _______________________________________________
>            gmx-users mailing list    gmx-users at gromacs.org
>         <mailto:gmx-users at gromacs.org>
>            <mailto:gmx-users at gromacs.org <mailto:gmx-users at gromacs.org>>
> 
>            http://www.gromacs.org/mailman/listinfo/gmx-users
>            Please search the archive at http://www.gromacs.org/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
>         <mailto:gmx-users-request at gromacs.org>
>            <mailto:gmx-users-request at gromacs.org
>         <mailto:gmx-users-request at gromacs.org>>.
> 
>            Can't post? Read http://www.gromacs.org/mailing_lists/users.php
> 
> 
> 
>         ------------------------------------------------------------------------
> 
> 
>         _______________________________________________
>         gmx-users mailing list    gmx-users at gromacs.org
>         <mailto:gmx-users at gromacs.org>
>         http://www.gromacs.org/mailman/listinfo/gmx-users
>         Please search the archive at http://www.gromacs.org/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
>         <mailto:gmx-users-request at gromacs.org>.
>         Can't post? Read http://www.gromacs.org/mailing_lists/users.php
> 
> 
>     -- 
>     ========================================
> 
>     Justin A. Lemkul
>     Graduate Research Assistant
>     Department of Biochemistry
>     Virginia Tech
>     Blacksburg, VA
>     jalemkul[at]vt.edu <http://vt.edu> | (540) 231-9080
>     http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin
> 
>     ========================================
>     _______________________________________________
>     gmx-users mailing list    gmx-users at gromacs.org
>     <mailto:gmx-users at gromacs.org>
>     http://www.gromacs.org/mailman/listinfo/gmx-users
>     Please search the archive at http://www.gromacs.org/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
>     <mailto:gmx-users-request at gromacs.org>.
>     Can't post? Read http://www.gromacs.org/mailing_lists/users.php
> 
> 

-- 
========================================

Justin A. Lemkul
Graduate Research Assistant
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================



More information about the gromacs.org_gmx-users mailing list