[gmx-users] Re: PME slows down

[Jason de Joannis]

> Date: Tue, 31 May 2005 21:58:03 +0200
> From: David <spoel at xray.bmc.uu.se>
> Subject: Re: [gmx-users] Re: gmx-users Digest, Vol 13, Issue 73
> To: Discussion list for GROMACS users <gmx-users at gromacs.org>
> Message-ID: <1117569483.22674.3.camel at localhost.localdomain>
> Content-Type: text/plain
>
> On Tue, 2005-05-31 at 14:21 -0400, Jason de Joannis wrote:
> >
> > > On Thu, 2005-05-26 at 11:51 -0400, Jason de Joannis wrote:
> > > > I am running a rather intensive (106K atoms) lipid bilayer simulation
> > > > with PME for 100 ns. Initially the time-to-completion estimate is
> > > > about 16 days. However over the course of 9 days the estimate increases
> > > > monotonically to 110 days. I have run quite a few simulations like this
> > > > before on the same hardware, though usually with cutoffs.
> > > Is it exploding?
> > NO.
> > >
> > > In parallel?
> > NO.
>
> Hard to say, check energy and density anyway.
Energy and density are normal.
> Which gromacs version? Is the machine out of memory?
v3.2, memory is fine.
> Are there other jobs running (run top).
nope.
> > >
> > > >        /Jason
> > > --
> > > David.
> > >

As I mentioned before, the only abnormality I can see are the constraint
warnings in the log file. Could it be that mdrun is spending more and more
time in constraint iteration loops?

 /Jason

-- 
Jason de Joannis, Ph.D.
Emory University
1515 Pierce Dr, Atlanta, GA 30329
Phone: (404) 712-2983
Email: jdejoan at emory.edu
http://userwww.service.emory.edu/~jdejoan