MPICH-GM + MKL trouble

Birukov Kirill <[email protected]> Wed, 11 Dec 2002 16:17:37 +0300
Newsgroups gmane.network.myrinet.general
Organization R&D Institute "Kvant"
Message-ID <[email protected]>
Hi!

I've got a problem running HPL (High Performance Linpack), compiled
with Intel MKL, under MPICH-GM. I am trying to run it on four 2-way
Xeon nodes, starting 1 process per node. I let MKL to parallelize on
each node (i.e. MKL creates 2 threads per node and both CPUs are
loaded). This trick works with MPICH from Argonne, which utilizes P4
device, but not MPICH-GM. Under MPICH-GM I've got some messages about
signal 11 after program start and it hangs. Has someone experience
with MPICH-GM and MKL?

P.S. When I limiting MKL with 1 thread per process, and running 2
processes per node, everything work fine. But I have information that
MKL performs better with automatic parallelization.

P.P.S. Configuration:
- 4 2-way Xeon nodes with 1 Gig of RAM.
- 1 Myrinet M3S-PCI64B adapter per node.
- RedHat Linux 7.3 (kernel 2.4.16 from MSC.Linux distro)
- GM 1.6.3
- MPICH for GM 1.2.4..8a
- MPICH from Argonne 1.2.4

--------------------------------------------------------------
|  Birukov Kirill                 [email protected]          |
|  Senior System Administrator                               |
|  R&D Institute Kvant            Moscow, Russia, 125438     |
|  Phone: +7 (095) 154-4362       4th Likhachevsky lane, 15  |
--------------------------------------------------------------