[Pw_forum] pw.x memory usage

Pierre.Cazade Pierre.Cazade at ul.ie
Tue Jul 8 09:53:48 CEST 2014


Hi Paolo,

Thank you for the information in particular for the item in the manual. I missed this one.

These tips are interesting as a poor choice in the input parameters may result in an unnecessarily large arrays. I wonder how to choose the cutoffs. By default, I used some parameters I found in the examples provided with QE but they may not be suitable for my system.

Thanks again for your help.

Regards,
Pierre 
________________________________________
From: pw_forum-bounces at pwscf.org [pw_forum-bounces at pwscf.org] on behalf of Paolo Giannozzi [paolo.giannozzi at uniud.it]
Sent: Monday, July 07, 2014 5:58 PM
To: PWSCF Forum
Subject: Re: [Pw_forum] pw.x memory usage

On Mon, 2014-07-07 at 15:27 +0100, Pierre-André Cazade wrote:

> This requires 97GB of RAM [...] it is normal that QE requires
> such a vast amount of memory

a 268-atom cell with some empty layers is not small. Consider that
the latest QE version stores by default all wavefunctions in memory,
since I/O is usually a more serious bottleneck than RAM. If you start
the code with "K_POINTS gamma", it should at least start and print out
some info on largest arrays. A few memory-saving tips are in the
user guide, item 5.0.0.10

P.
--
 Paolo Giannozzi, Dept. Chemistry&Physics&Environment,
 Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
 Phone +39-0432-558216, fax +39-0432-558222

_______________________________________________
Pw_forum mailing list
Pw_forum at pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum



More information about the users mailing list