[QE-users] Fwd: projwfc.x out of memory

Paolo Giannozzi p.giannozzi at gmail.com
Wed Mar 28 11:47:44 CEST 2018


As far as I know, projwfc.x does not need more memory than a scf or nscf
calculation (It should actually take somewhat less). In a similar recent
case,
neither I nor other people in Cineca found any evidence of memory leaks or
sources of unexpected memory usage. WIthout more information, it is hard to
say more (and even with more information: large jobs are not easy to debug)

Paolo

On Wed, Mar 28, 2018 at 3:21 AM, Aldo Ugolotti <a.ugolotti at campus.unimib.it>
wrote:

> Dear QE Users,
>
> I am recently encountering an error running projwfc.x. The nscf
> calculation to increase the k points mesh runs smoothly, however few
> minutes after starting the projwfc.x task, the process is killed by the
> system with an "out of memory" message. This happens on a large cluster
> with 180 Gb di memory per node. This is not resolved if I reduce the
> energy window into which the dos is calculated.
>
> Is there a way to resolve the issue?
>
> Regards,
>
> --
> Aldo Ugolotti
>
> PhD student
> Materials Science Dept. U5,
> Università degli Studi di Milano-Bicocca
> via Cozzi 55,
> 20125 Milano (MI)
> Italy
> e-mail: a.ugolotti at campus.unimib.it
>
> _______________________________________________
> users mailing list
> users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users




-- 
Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche,
Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20180328/a6ab0a14/attachment.html>


More information about the users mailing list