Sure, LAM will work fine in this configuration.
If you run all your processes on the same node, LAM will use shared
memory for its communication, and it should therefore be quite fast.
On Jan 18, 2005, at 4:07 AM, Noel Jakse wrote:
> Hello,
>
> I'm very interested in LAM/MPI since I built mpi fortran77 code for
> molecular dynamics simulations actually running under IBM SP3 power4
> in a french computer ressource center.
>
> Besides, my lab has bought a server with bi-processor motherboard and
> 2 Intel Xeon 2.8GHz and 2Gbyte RAM, working properly under linux
> mandrake 10.1 official. I would like to ask you if LAM/MPI will work
> on such a server with node=1 and cpu=4, and if you know somebody who
> manage to work with LAM/MPI in a similar configuration ?
>
> Thank you to the lam/mpi team in advance
> All the best,
>
> Noel Jakse
>
> Laboratoire de Theorie de La Matiere Condensee
> Universite de Metz
> 1, boulevard Arago
> 57078 Metz Cedex 3
> Tel : + 33 (0)3 87 31 58 50
> fax : + 33 (0)3 87 31 58 01
> email : jakse_at_[hidden]
> _______________________________________________
> This list is archived at http://www.lam-mpi.org/MailArchives/lam/
>
--
{+} Jeff Squyres
{+} jsquyres_at_[hidden]
{+} http://www.lam-mpi.org/
|
