You might also want to use TCP or any of the other C2C RPIs -- lamd is
definitely slower than the C2C versions.
lamd mode should not be cause *hours* of slowness, but it's worth
checking.
Also, note that in an all-Linux environment, you should probably use the
"-O" switch to mpirun to ensure that LAM thinks you're running on a
homogeneous system and therefore will not do endian byte swapping.
On Wed, 9 Apr 2003, Venkat Ramanan K wrote:
> Hi,
>
> We are a bunch of undergraduate students and have put up a cluster
> of 16 IBM's using LAM for the purpose of running GROMACS. During
> our simulation test runs we noticed that the running time
> increases as the number of nodes in the LAM topology increases. We
> couldn't diagonise the problem...plz help.
>
> We used the folowing commands:
>
> mpirun N -lamd -nger -nsigs mdrun -v -s full........-g flog
>
> We got the following running times...
>
> 1 node - 13hrs 5:27
> 2 nodes - 16hrs 12:59
> 4 nodes - 17hrs 32:14
> 8 nodes - 45hrs 00:36
> 16 nodes - 45hrs 46:55
>
> Venkat
>
> _______________________________________________________________________
> Odomos - the only mosquito protection outside 4 walls -
> Click here to know more!
> http://r.rediff.com/r?http://clients.rediff.com/odomos/Odomos.htm&&odomos&&wn
>
> _______________________________________________
> This list is archived at http://www.lam-mpi.org/MailArchives/lam/
>
--
{+} Jeff Squyres
{+} jsquyres_at_[hidden]
{+} http://www.lam-mpi.org/
|
