[DFTB-Plus-User] How many atoms DFTB+ can handle

[Bálint Aradi]

> 
> I can run my jobs on a cluster which has 200 GB RAM and with 64
> processors (AMD).

I guess, you should be able to run a few thousand atoms on that. For
let's say 4000 atoms, I'd expect one total energy calculation (ca. 10
SCC-cycles + force calculation) taking less than an hour. Of course,
exact timings are depending on your actual system (how many orbitals),
and on your hardware.

> If possible, can anybody give the information on the speed of the
> calculations with a change in the number of processors and/or RAM size.

I attach one timing data for the MPI-branch. It was made on a rather
"old" machine. But, I strongly recommend to do such scaling tests on
yourself on your specific architecture. Especially, as you have an
SMP-machine, you should also try the normal version of DFTB+ linked with
some threaded LAPACK-library.

  Best regards,

  Bálint

-- 
Dr. Bálint Aradi
Bremen Center for Computational Materials Science, University of Bremen
http://www.bccms.uni-bremen.de/cms/people/b_aradi/

-------------- next part --------------
A non-text attachment was scrubbed...
Name: timing.pdf
Type: application/pdf
Size: 9478 bytes
Desc: not available
URL: <http://www.dftb-plus.info/pipermail/dftb-plus-user/attachments/20140505/220a4167/attachment-0001.pdf>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 263 bytes
Desc: OpenPGP digital signature
URL: <http://www.dftb-plus.info/pipermail/dftb-plus-user/attachments/20140505/220a4167/attachment-0001.sig>