Many thanks for all the valuable inputs. But, now I am little confused
about using ntt=1 or ntt=3 for microsecond long production run. As
suggested by Andrew Schaub ( and Ross,
http://archive.ambermd.org/201103/0431.html), using ntt=1 can avoid NANs in
long run which come with ntt=3. But, as discussed here, using ntt=3 and
ig=-1 will take care about the Langevin synchronization effect. Is it safe
to go with
http://archive.ambermd.org/201103/0431.html for long run?
Thanks in advance
Anu
On Thu, Apr 30, 2015 at 12:48 PM, David A Case <case.biomaps.rutgers.edu>
wrote:
> On Wed, Apr 29, 2015, anu chandra wrote:
>
> > Many thanks Jason and Prof. Case. A quick query here is about the value
> of
> > 'gamma_In'. Is it safe to set a value of 5 ps^-1 for gamma_In? Is it
> going
> > to effect the computational time?
>
> The magnitude of gamma_ln has no effect on computation time. Values in the
> range of 1 to 5 ps^-1 are very commonly used. Values outside that range
> may
> have merit in special cases, but you should consider that a research topic.
>
> ...dac
>
>
> _______________________________________________
> 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 Thu Apr 30 2015 - 05:30:04 PDT
