[CPMD-list] About CPMD parallel

Axel Kohlmeyer akohlmey at vitae.cmm.upenn.edu
Sun Jun 18 07:57:28 CEST 2006


On Sun, 18 Jun 2006, Donghee Lim wrote:

DL> Dear All
DL> 
DL> I have some problems in the use of CPMD parallel in our clusters that 
DL> include 4 CPUs.
DL> The following command was used for the CPMD parallel calculation;
DL> 
DL> >> mprun -l "diamond.engin.umich.edu  2" 
DL> >> /l/software/cpmd/bin/cpmd.mpi.x input.inp > output.out

please state what software exactly you are using and what
platform you are running on.

DL> I thought the above command would make my calculation run in two CPUs 
DL> at the same time, but as shown the following output file, I guess just 
DL> one CPU was used (NCPU: 0).

yes, unless you are running with the gromos/cpmd qm/mm interface code...


DL> PARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARA
DL>   NCPU     NGW     NHG  PLANES  GXRAYS  HXRAYS ORBITALS Z-PLANES
DL>      0  192294 2149751     216    6385   31949      14       1
DL>                 G=0 COMPONENT ON PROCESSOR :     0
DL> PARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARA
DL> 
DL> Do I need to assign any commands to the input file for the use of CPMD 
DL> parallel?

no. CPMD gets the information from the MPI-API, so this is purely an
issue of your MPI implementation. please check out the documentation,
ask your system administrator or ask the user support of that package
(looks like scali manage, but this is just a wild guess).

DL> Also, what does the number 2 in my command line stand for? (I guess it 
DL> stands for 2 CPUs)..

how should we know? the MPI standard does not specify how a parallel 
job has to be started (except that the frontend program should be 
named 'mpirun'. so you have to refer to the MPI documentation. best
you check out some of the standard MPI test/benchmark examples, before 
you move to CPMD on a platform you are not familiar with.


cheers,
    axel.
DL> Thank you,
DL> Donghee Lim
DL> 
DL> 
DL> 
DL> _______________________________________________
DL> CPMD-list mailing list
DL> CPMD-list at cpmd.org
DL> http://cpmd.org/mailman/listinfo/cpmd-list
DL> 

-- 
=======================================================================
Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.




More information about the CPMD-list mailing list