[MITgcm-devel] linux_ia64_ifort+mpi_altix_gfdl

[Patrick Heimbach]

Dimitris,

testing this currently as possible cause
why version3 ECCO doesn't run at GFDL
(but identically transfered code to NAS)
runs at NAS with _nas optfile.

-p.

On Wed, 2006-12-06 at 08:00 -0800, Dimitris Menemenlis wrote:
> Baylor, thank you for info.  Do you, or anyone else on devel list, have 
> experience with removing the "-mp" flag, i.e., does all hell break loose?
> 
> I am thinking of experimenting with following options file on the NAS Altix:
> http://mitgcm.org/cgi-bin/viewcvs.cgi/MITgcm_contrib/high_res_cube/code-mods/linux_ia64_ifort+mpi_altix_nas
> 
> > I did a little testing on a medium sized simulation (64 processors) with -ip
> > and -ipo.  I saw modest improvements in speed (couple of percent) with -ip
> > and no further increase with -ipo.  Of course, -ipo takes more than twice as
> > long to compile.  But, I think there is no other loss in using -ipo if you
> > like...
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel
-- 
Patrick Heimbach | heimbach at mit.edu | http://www.mit.edu/~heimbach
MIT | EAPS 54-1518 | 77 Massachusetts Ave | Cambridge MA 02139 USA
FON +1-617-253-5259 | FAX +1-617-253-4464 | SKYPE patrick.heimbach