Re: [AMBER-Developers] IPS in pmemd

From: Romelia Salomon <romelia.caltech.edu>
Date: Fri, 20 Jul 2012 07:43:37 -0700

Hi Dave,

I'll see if i can add ips =2 cleanly and i'll let you know. I'll also check
couple things for many procs because Ross and i agree with your concerns.

Have a nice day!

--
************************************
Romelia Salomon
Walker Group
398 San Diego Supercomputer Center
UC San Diego
-----Original message-----
From: David A Case <case.biomaps.rutgers.edu>
To: AMBER Developers Mailing List <amber-developers.ambermd.org>
Sent: Fri, Jul 20, 2012 05:01:21 PDT
Subject: Re: [AMBER-Developers] IPS in pmemd
On Thu, Jul 19, 2012, Romelia Salomon wrote:
> 
> As Ross, I also don't think having ips for VdW should be a problem,
Just note that IPS=1 includes both electrostatic and vdW as IPS; IPS=2
has just electrostatics, and should be faster, but I don't know by how much.
> 
>   #procs NO-longrange    PME-orig  PME-new  IPS-new
>     4      409.21         553.72    550.47   475.03
>     8      219.93         290.28    291.92   250.88
>    16      127.92         159.21    159.51   144.96
>    32       75.23          91.02     90.19    79.76
>    64       73.58          79.80     79.22    73.35
> 
Thanks for the update!  I only see problems beyond 64 threads, so I'll keep
investigating.
....dac
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers

Received on Fri Jul 20 2012 - 08:00:03 PDT