[ADF-LIST] Reaxff performance, system size vs. core

Alexei Yakovlev yakovlev at scm.com
Mon Sep 25 17:07:38 CEST 2017

Dear Praveen,


On 21/09/2017 08:26, Praveen Kumar wrote:
> Hi,
>
> I want to know the performance of Reaxff in ADF. How many atoms can I 
> choose for Reaxff simulation (Maximum limit) and what will be the 
> required number of cores?
These are dictated by the current hard limit on the number of atoms per 
core: there may be not more 100000 atoms per core, including ghost 
atoms. Besides, for good parallel performance there should be at least 
1000 atoms per core. It can be more or less than that depending on which 
interactions take the most time to compute for the given system.

Kind regards,
Alexei
> How much time will this job take to finish?
>
> Does anybody have this performance graph? Any suggestion will be helpful.
>
>
>
>
> Regards
> PRAVEEN KUMAR
> Research scholar
> INDIAN INSTITUTE OF SCIENCE
> EDUCATION AND RESEARCH PUNE, India
>
>
> _______________________________________________
> ADFlist mailing list
> ADFlist at scm.com
> https://lists.scm.com/mailman/listinfo/adflist

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.scm.com/pipermail/adflist/attachments/20170925/4d727ca6/attachment-0001.html>


More information about the ADFlist mailing list