AMBER Archive (2009)

Subject: RE: [AMBER] pmemd vs sander.MPI

From: Ross Walker (ross_at_rosswalker.co.uk)
Date: Wed Nov 04 2009 - 12:33:23 CST


Hi Taufi,

> I am using sander.MPI to run a dynamics simulation with a restraint
> mask
>
> Molecular Dynamics for 2ns - only Surface atoms are fixed
>
> can i run the above with pmemd by changing the last two lines
> Molecular Dynamics for 2ns - only Surface atoms are fixed

Yes. PMEMD just needs the old restraint syntax. There is a tool called
ambmask in $AMBERHOME/exe/ which will convert an AMBER mask into the old
group format for you.
 
> Also, i do remember talking with Dr. Walker about pmemd being faster
> and
> just as accurate as sander.MPI, is that also true when using 32 or 64
> processors, or is it especially true when using large number of
> processors?

It will be a LOT faster. If you are running simulations within PMEMD's
feature set you should probably always use it since it will be faster even
on 1 processor. On 32+ it will make a big difference. You should be seeing
performance equivalent to the following (this was run with pmemd)

http://teragrid.org/userinfo/aus/amber_benchmark.php

All the best
Ross

/\
\/
|\oss Walker

| Assistant Research Professor |
| San Diego Supercomputer Center |
| Tel: +1 858 822 0854 | EMail:- ross_at_rosswalker.co.uk |
| http://www.rosswalker.co.uk | PGP Key available on request |

Note: Electronic Mail is not secure, has no guarantee of delivery, may not
be read every day, and should not be used for urgent or sensitive issues.

_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber