[MITgcm-devel] a 6.cpu output for fizhi-cs-32x32x10

[Jean-Michel Campin]

Hi Andrea,

I checked-in an other output file (in addition to output.txt)
for the fizhi-cs-32x32x10 experiment, that was produced on 6
cpu (pgf_mpi, -ieee, myrinet cluster).

The reason is that I think this 6.cpu result is more
reliable, and is closer to what you are really running.
(1 tile / processor).

I found a problem with landfrac that gives wrong things
in the case where there is more than 1 tile / processor
(I will check-in this fix), but there might be other pb
like this, since I cannot get a good agreement between
1 and 6 processors (even after 1 iter, found differences
~ 1.K for theta at the 1rst dynamical level).

See you,

Jean-Michel.