[MITgcm-support] AIX and numerical diffusion

[Haferman, Jeffrey (Jeff) (CIV)]

Got it!!!!

The key was to use the "sp4+mpi_datastar" file with genmake2 (even
though I'm not using the SDSC datastar machine).  I had been using the
sp4+mpi file (and commented out the references to libraries that I
wasn't using); though I'm not sure exactly what was causing the
divergent behavior, NOW my IBM runs converge and the output matches
*exactly* with my Linux runs....

Jeff



-----Original Message-----
From: mitgcm-support-bounces at mitgcm.org
[mailto:mitgcm-support-bounces at mitgcm.org] On Behalf Of Haferman,
Jeffrey (Jeff) (CIV)
Sent: Thursday, April 26, 2007 3:58 PM
To: mitgcm-support at mitgcm.org
Subject: RE: [MITgcm-support] AIX and numerical diffusion

I rebuilt using genmake2 with the -ieee flag (which basically adds the
-qstrict option to the compilation) and though the output values
changed, I still got the same divergent behavior in the end....

I will play with some different compilation flags (remove all
optimization to begin with) and see if I can fix the problem.  If you
have any other ideas, please let me know.

Thank you,
Jeff


------------------------------------------------------------------------
--------------------------------------------
> I guess I can try to decrease the timestep on the IBM to see if it
> helps, but does has anyone else had this type of experience when
moving
> a model to a new platform, and can you offer some suggestions?  I am
> using the IBM xlf and xlc compilers under their POE environment, if
that
> makes any difference.

Please rerun genmake2 with the -ieee flag, rebuild and run and let us
know if 
you experience the same problem. We may need to change the default 
optimization level on the IBMs.

Constantinos

_______________________________________________
MITgcm-support mailing list
MITgcm-support at mitgcm.org
http://mitgcm.org/mailman/listinfo/mitgcm-support