Hi Neha,
To clarify this a little further. In the case of pmemd.cuda the cut off
value for GB calculations is actually ignored so there really is no
difference between 10, 99, 999 or 9999. The code simply uses it as an
initial check to make sure the cutoff is set to a value larger than the
longest dimension of the system. This is purely to check if you are
expecting the simulation to be run without a cutoff or not.
For the CPU code the cutoff is used but setting it to a value greater than
your system size ever reaches means you will run without a cutoff. So
generally 999 or 9999 will be good since your system is unlikely to be
longer than 999 angstroms.
All the best
Ross
> -----Original Message-----
> From: Jason Swails [mailto:jason.swails.gmail.com]
> Sent: Tuesday, November 01, 2011 7:10 AM
> To: AMBER Mailing List
> Subject: Re: [AMBER] cutoff in amber cuda
>
> If your system is smaller than 999 Angstroms in diameter, then there is
> no difference. rgbmax is a different parameter altogether, and
> specifies the cutoff for GB radius calculation.
>
> Note that 999 angstroms is very large and is effectively infinite for
> all systems. Does anyone have an example of where I'm wrong?
>
> All the best,
> Jason
>
> --
> Jason M. Swails
> Quantum Theory Project,
> University of Florida
> Ph.D. Candidate
> 352-392-4032
>
> On Oct 31, 2011, at 10:23 PM, Neha Gandhi <n.gandhiau.gmail.com> wrote:
>
> > Dear Amberusers,
> >
> > I am running implicit solvent simulations (GB) using pmemd.cuda.MPI.
> Is
> > there any difference using cut=999 (amber tutorials) and cut=9999 in
> > combination with rgbmax (amber cuda benchmark test)?
> >
> > Thanks,
> > Neha
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Nov 01 2011 - 09:00:04 PDT
