[gmx-developers] pme in release-4-6 with domain decomposition

[Dommert Florian]

On Mon, 2012-04-09 at 10:45 -0400, Roland Schulz wrote:
> Hi,
> 
> 
> we need a bit more detail. What architecture (/what nb kernel)? What
> compiler? 
> 
> 

Hi,

the architecture is X86_64 and the compiler is gcc4.4 shipped with
Ubuntu 10.4. OpenMPI1.4.3 compiled with gcc4.4 has been used to build
Gromacs in double precision. The error arises with the NB_GENERIC and
the standard ASM kernels for x86_64. 


Cheers,

Flo

> Roland
> 
> On Mon, Apr 9, 2012 at 8:54 AM, Dommert Florian
> <dommert at icp.uni-stuttgart.de> wrote:
>         Hi,
>         
>         today I tested the new release-4-6 branch, but inconsistent
>         electrostatic forces arise.
>         
>         As test system I used the spc216 water box and calculated the
>         forces
>         with ewald and pme. If only a single processor is used the
>         forces are
>         consistent. Also for multiple processes and particle
>         decomposition the
>         forces are correct. However if domain decomposition used, the
>         forces are
>         sometimes in the order of 10^102 (correct average maximum
>         force 2402
>         kJ/mol/nm).
>         
>         If the test systems consists of 500 positive and negative
>         atomic ions or
>         molecular ions, everything is fine, too, indepenent of the
>         decomposition
>         scheme.
>         
>         Only the model systems with water make trouble (I tried SPC/E
>         and
>         TIP-4P).
>         Does anyone has an idea, what the reason for this problem
>         might be ?
>         
>         Cheers,
>         Flo
>         
>         
>         
>         --
>         Florian Dommert
>         Dipl. - Phys.
>         
>         Institute for Computational Physics
>         University Stuttgart
>         
>         Pfaffenwaldring 27
>         70569 Stuttgart
>         
>         EMail: dommert at icp.uni-stuttgart.de
>         Homepage: http://www.icp.uni-stuttgart.de/~icp/Florian_Dommert
>         
>         Tel.: +49 - (0)711 - 68563613
>         Fax.: +49 - (0)711 - 68563658 
> 
> 
> 
> 
> -- 
> ORNL/UT Center for Molecular Biophysics cmb.ornl.gov
> 865-241-1537, ORNL PO BOX 2008 MS6309
> 
> -- 
> gmx-developers mailing list
> gmx-developers at gromacs.org
> http://lists.gromacs.org/mailman/listinfo/gmx-developers
> Please don't post (un)subscribe requests to the list. Use the 
> www interface or send it to gmx-developers-request at gromacs.org.

-- 
Florian Dommert
Dipl. - Phys.

Institute for Computational Physics
University Stuttgart

Pfaffenwaldring 27
70569 Stuttgart

EMail: dommert at icp.uni-stuttgart.de
Homepage: http://www.icp.uni-stuttgart.de/~icp/Florian_Dommert

Tel.: +49 - (0)711 - 68563613
Fax.: +49 - (0)711 - 68563658
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-developers/attachments/20120409/5e8569b9/attachment.sig>