On Thu, Feb 23, 2012, Volker Lesch wrote:
>
> I have a big problem using Sander. I want to start a MD-Simulation with
> ipol=1 and ew_type=1. This combination produces an error like "Floating
> point exception". I use different inp.crd and prmtop files, but it
> crashes every time. If I use ipol=0 oder ew_type=0, then it works fine
> and I get no errors. I want to use the Ewald summation because I want to
> reproduce data which are produced with these parameters. Is this kind of
> error known? Thanks for your help.
I'm guessing that no one has tried this combination in a long time. [For
developers: what is presumably the same error can be obtained by going
to $AMBERHOME/test/dna_pol, and setting ew_type=1 in the Run.dna_pol script.]
I'm hoping that running under a debugger will give some useful clues.
Note that in general, PME is more accurate than regular Ewald, unless you
have a very high k-limit on the Ewald sum. You can also check the accuracy of
the PME result by using a finer spacing.
Thanks for the report....dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Feb 23 2012 - 09:00:01 PST
