[DFTB-Plus-User] OpenMP parallelistaion

[Balint Aradi]

Dear Donglai,

>     I have downloaded the source code and the new version of the manual. Thanks so much for your effort.
>     I found that in the Chaper 1 of the manual, there is a word "OpenMP parallelisation", what does it mean?
>     I did not find any content about the parallelization in the manual.

  The OpenMP parallelisation means, that the code contains OpenMP
directives. Using a compiler supporting OpenMP (e.g. Intel, Lahey) and
setting the appropriate compiler options (e.g. -mp or similar), you can
obtain a code, which is able to execute some of its tasks in multiple
threads parallel (on a multi processor or multi core machine with shared
memory). You have eventually set extra environment variables when
running the code in order to enable parallel execution (e.g.
OMP_NUM_THREADS). You should consult the manual of your compiler for the
details.

  Please note, that for big systems DFTB+ spends about 90% of its
execution time in the LAPACK diagonalisation routine. Therefore, you
should make sure, that you're linking the code with an OpenMP capable
BLAS/LAPACK library (e.g. MKL).

  Best regards,

      Bálint


-- 
========================================================================
Bálint ARADI, Ph.D. <aradi at bccms.uni-bremen.de>
Bremen Center for Computational Materials Science, University of Bremen
Room 3.10, Building TAB (A), Am Fallturm 1, 28359 Bremen, Germany
Tel.: +49 421 218 7421, Fax: +49 421 218 4764
Web: http://www.bccms.uni-bremen.de/de/mitarbeiter/b_aradi/
========================================================================

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 252 bytes
Desc: OpenPGP digital signature
URL: <http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20070924/a0800db5/attachment.sig>