[gmx-users] Haphazard Results with LAM-MPI

[Anton Feenstra]

David wrote:
> On Wed, 2003-04-09 at 06:13, 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 
>>
[...]
>
> Check the results on the benchmark web page. We have results for a
> switched network and as long as you have a reasonably large system (and
> no PME) it scales half decent at least.

I have similarly (bad) results with our cluster in Amsterdam,
we suspect some of the settings of LAM or mpich (we have both
installed) are not optimal, but since changing and testing it
is pretty tedious, and for the time we have about as many
different simulations to run as we have CPU's, I have not yet
taken the time to get to the bottom of it.

I'd be interested to know what hardware you are running on,
CPU, main board, ethernet card, and which software, linux
version, and specific kernel options and module versions too.
Maybe we can find a common denominator to be the problem here.


-- 
Groetjes,

Anton
  _____________ _______________________________________________________
|             |                                                       |
|  _   _  ___,| K. Anton Feenstra                                     |
| / \ / \'| | | Dept. of Pharmacochem. - Vrije Universiteit Amsterdam |
|(   |   )| | | De Boelelaan 1083 - 1081 HV Amsterdam - Netherlands   |
| \_/ \_/ | | | Tel: +31 20 44 47608 - Fax: +31 20 44 47610           |
|             | Feenstra at chem.vu.nl - www.chem.vu.nl/~feenstra/       |
|             | "If You See Me Getting High, Knock Me Down"           |
|             | (Red Hot Chili Peppers)                               |
|_____________|_______________________________________________________|