I'm using the 8.0 and 9.0 intel compilers on a Xeon cluster without
any problems. And I didn't have to specify any configure flags. Just
the default one:
./configure --prefix=/usr/local/ompi/1.0.2 --with-mvapi=/usr/local/
topspin/ --enable-mpi-f90 --enable-mpi-f77 --enable-mpi-cxx --enable-
mpi-profiling --enable-picky --with-f90-max-array-dim=2 F77=/opt/
intel/fce/9.0/bin/ifort F90=/opt/intel/fce/9.0/bin/ifort CC=/opt/
intel/cce/9.0/bin/icc CXX=/opt/intel/cce/9.0/bin/icpc
Thanks,
george.
On Apr 12, 2006, at 6:03 PM, Jeffrey B. Layton wrote:
Troy Telford wrote:
On Wed, 12 Apr 2006 14:34:06 -0600, Jeffrey B. Layton
<layto...@charter.net> wrote:
Good afternoon,
While we're on the subject of building OpenMPI-1.0.2 with
Intel, I'm having trouble building OpenMPI-1.0.2 with Intel 9.0.
I'm starting to wonder if I'm doing something magical; I've compiled
Open MPI 1.0.2 with Intel 9.0 (EM64T version) with no problems. I do
use slightly different configure options, but you disable a few
things
I don't.
I can build it fine on the Opterons but some of the NPB codes fail
when running.
On the Xeons, I can't seem to get it to build. I'll retry with your
configure options but I'll disable ROMIO.
Thanks!
Jeff
_______________________________________________
users mailing list
us...@open-mpi.org
http://www.open-mpi.org/mailman/listinfo.cgi/users
