[gmx-users] IRIX PROBLEMS

[David van der Spoel]

On Tue, 2003-12-09 at 10:44, Raj Badhan wrote:
> Dear Users,
> Firstly thank you too all who have passed on suggestions for my 
> EM runs with ATP, I'm half way there now, but I'm sure I will be 
> bugging you all again for more assistance.
> Secondly, we are having troubles running our simulations on the 
> Universities high performance computer system.
> They have installed Gromacs in double precision for us, and I can 
> quite happily run my simulations using a single processor (-np 1) 
> by issuing the MPIRUN command. But when we try to run in 
> parallel on 4 processors, we get an error message back stating 
> that the run input files were designed 256 nodes and were 
> expecting to be for 4 nodes.

please start your tests with a simple system, e.g. the water box from
the tutorial
then:

grompp -np 4
mpirun -c 4 mdrun

should work (for LAM, for SGI mpirun you may need different command
line).

Energy minimization in parallel does not work IIRC.

> This occurs when initiating the MDRUN command for the first set of 
> energy minimisations on out protein.
> We are not allowed interactive access to the server, and must 
> submit batch jobs of each commands. Now, these commands 
> seem to be fine as we can run the sims. on a single processor, but 
> there is always a problem when running in parallel.
> Could this be a conflict with some of the libraries, something like 
> FFTW and LAM-MPI??
The error message you showed suggests otherwise.

-- 
David.
________________________________________________________________________
David van der Spoel, PhD, Assist. Prof., Molecular Biophysics group,
Dept. of Cell and Molecular Biology, Uppsala University.
Husargatan 3, Box 596,  	75124 Uppsala, Sweden
phone:	46 18 471 4205		fax: 46 18 511 755
spoel at xray.bmc.uu.se	spoel at gromacs.org   http://xray.bmc.uu.se/~spoel
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++