[petsc-users] Strange efficiency in PETSc-dev using OpenMP

[Jed Brown]

Barry Smith <bsmith at mcs.anl.gov> writes:
>    So when you compile the software to use OpenMP it is slow;
>    regardless of whether you use OpenMP explicitly or not. When you
>    compile the software to NOT use OpenMP then it is much faster?

Are you running multiple MPI processes?  How are the affinities set?
Are any other jobs competing for the same pinned core as your
application?
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 835 bytes
Desc: not available
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20130923/99ae4a23/attachment.pgp>