Re: [AMBER] Possible bug in PMEMD (appearence of synchronization artifacts between parallel tasks)

From: Adrian Roitberg via AMBER <amber.ambermd.org>
Date: Fri, 29 Mar 2024 11:09:37 -0400

Hi Charo

Could you share prmtop/input/rst/submission script so I can try it locally ?

Thanks


On 3/28/24 10:11 AM, Charo del Genio via AMBER wrote:
> [External Email]
>
> Dear all,
>        continuing from my previous email, this (attached) is the
> temperature of the system, from the mdout files. As before, the first
> 200 ns are on 128 processors, and the last 50 ns are on 8 processors.
> Repeating the run with sander, I do not see this behaviour (so far).
>
> Any ideas on how to proceed would be much appreciated. At this stage,
> it seems to me like results obtained with ntt=3 and pmemd.MPI could be
> not reliable, at least in some cases.
>
>
> Cheers,
>
> Charo
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber

-- 
Dr. Adrian E. Roitberg
V.T. and Louise Jackson Professor in Chemistry
Department of Chemistry
University of Florida
roitberg.ufl.edu
352-392-6972
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Mar 29 2024 - 08:30:03 PDT
Custom Search