[DFTB-Plus-User] Memory usage and CPU time

[Balint Aradi]

Dear Zbigniew Romanowski,

> I would like to apply DFTB+ software to study the SiC clusters (without 
> periodic boundary conditions) with 10^3 or 10^4 atoms. However, I do not 
> have access to this software and I cannot estimate the memory usage and 
> CPU time for this kind of job.
>  
> Could anyone provide me the memory usage and CPU time for 1 CPU machine 
> for system with 10^4 atoms?

  For SiC clusters, you should be able to calculate 1000 atoms with a 
memory requirement of about 1 GB. Memory usage scales approx. 
quadratically with the number of orbitals at the moment. About timings 
for various systems, see

B. Aradi, B. Hourahine and T. Frauenheim.
DFTB+, a Sparse Matrix-Based Implementation of the DFTB Method.
J. Phys. Chem. A, 111:5678, 2007.

http://dx.doi.org/10.1021/jp070186p

   Best regards,

    Bálint


Ps. Ben Hourahine is developing an order N version of the code, which 
has a much better scaling on memory consumption and execution times, and 
would be probably more suitable for the system sizes you're envisioning. 
As this extension had not been officially released yet, you should 
contact him directly for further details.

-- 
Dr. Bálint Aradi
Bremen Center for Computational Materials Science, University of Bremen
TAB A/3.10, Am Fallturm 1, 28359 Bremen, Germany, Tel.: +49 421 2187421
http://www.bccms.uni-bremen.de/en/people/home/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://www.dftb-plus.info/pipermail/dftb-plus-user/attachments/20080521/c3fb9c58/attachment-0001.pgp