[ase-users] Problems with LAMMPSlib interface for >8 CPUs

Andrew Logsdail LogsdailA at cardiff.ac.uk
Sat May 11 16:23:50 CEST 2019


Hi Noam,

I’m with you on this, I can’t really see what the cause would be! Our nodes are 40 cores, so 8 -> 16 does not invoke inter-node communication, nor changes in filesystem.

At current my thought is it may be a specific error to my build, rather than generic to LAMMPSlib, hence why I’m hoping someone can run comparable testruns.

Thanks in advance for any further thoughts available!

Andy

On 11 May 2019, at 15:20, Noam Bernstein <noam.bernstein at nrl.navy.mil<mailto:noam.bernstein at nrl.navy.mil>> wrote:

On May 11, 2019, at 10:17 AM, Noam Bernstein <noam.bernstein at nrl.navy.mil<mailto:noam.bernstein at nrl.navy.mil>> wrote:

Well, since I'm at least partially responsible for the starting version for Ellad's work (which we improved in our python nested sampling code pymatnest), maybe I can take a look.


Starting along this direction, I find it hard to imagine where in LAMMPSlib things would be dependent in any explicit way on the number of MPI tasks.  So, is there any other way that 8 is different from 16 on your system?  Is that the point at which you go from 1 node to 2, and they have separate filesystems, for example?

Noam


____________
|
|
|
U.S. NAVAL
|
|
_RESEARCH_
|
LABORATORY

Noam Bernstein, Ph.D.
Center for Materials Physics and Technology
U.S. Naval Research Laboratory
T +1 202 404 8628  F +1 202 404 7546
https://www.nrl.navy.mil<https://eur03.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.nrl.navy.mil&data=01%7C01%7CLogsdailA%40cardiff.ac.uk%7C68ef9baee3ff4a0717d308d6d61bd0cc%7Cbdb74b3095684856bdbf06759778fcbc%7C1&sdata=f0sPlDXEiQTD6Y7eIDUy2IHYmEdOL8qJHcY0cMTwdXU%3D&reserved=0>



—

Dr. Andrew Logsdail

School of Chemistry,
Cardiff University,
Main Building,
Park Pl,
Cardiff CF10 3AT

T: +44 2922 510 162
E: LogsdailA at cardiff.ac.uk<mailto:LogsdailA at cardiff.ac.uk>
W: http://logsdail.github.io
@: http://www.twitter.com/a_logsdail

Yr Ysgol Cemeg,
Prifysgol Caerdydd,
Y Prif Adeilad,
Plas-y-Parc,
Caerdydd,
CF10 3AT

F: +44 2922 510 162
E: LogsdailA at cardiff.ac.uk<mailto:LogsdailA at cardiff.ac.uk>
W: http://logsdail.github.io
@: http://www.twitter.com/a_logsdail

The University welcomes correspondence in Welsh or English.
Mae'r Brifysgol yn croesawu gohebiaeth yn Gymraeg neu'n Saesneg.



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.fysik.dtu.dk/pipermail/ase-users/attachments/20190511/3cdc926f/attachment-0001.html>


More information about the ase-users mailing list