Dear Jason,
Thank you very much for answering so quickly.
We should have mentioned that the problem only appears when we run EVB
simulations. A run with a single EVB state works fine on multiple cores.
We have exactly 69 atom types in the first VB state, and 4 more types in the
second VB state, so a total of 73 atom types.
We did not try to run PMEMD because our understanding was that EVB is not
implemented.
We are going to recompile sander with the debug flag as you suggested, and
we will let you know the results as soon as we get them.
Thank you again,
Elise Duboué-Dijon
> On Thu, Feb 28, 2013 at 9:24 AM, Elise Duboué-Dijon <
> elise.duboue-dijon.ens.fr> wrote:
>
>> Dear Jason,
>>
>> Following your advices, we changed the amber12 code to be able to run it
>> with systems that have more than 60 atom types.
>>
>
> How many atom types do you have _exactly_? The changes you made below
> correspond to just 100 atom types...
>
> Also, can you use pmemd instead of sander? pmemd does not suffer from
> this
> limitation. (As a note, I implemented the changes in the RCW: WARNING
> comment block at the top of parms.F90 for the upcoming Amber version).
>
> Can you re-compile with the "-g" flag to add debug symbols? (You should
> be
> able to configure with the -debug flag), and re-run. Hopefully that will
> provide a backtrace so we know where the segfault is occurring.
>
> Thanks,
> Jason
>
> --
> Jason M. Swails
> Quantum Theory Project,
> University of Florida
> Ph.D. Candidate
> 352-392-4032
> _______________________________________________
> 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 Feb 28 2013 - 09:00:03 PST
