[CSC 435] Clarification...

[Andrew J. Pounds]

Since you have to do multiple runs of different matrix dimensions, it is 
also possible to set up something like this if you set up your fortran 
to read in the matrix dimenstion.  This is preferred over running 10 
jobs, each with a 10000x10000 grid.

#PBS -N JACOBI_TEST
#PBS -j oe
#PBS -m abe
#PBS -l nodes=1:hammer:ppn=42
#PBS -l walltime=1:00:00
#
cd ./midtermprob
echo 1000 >param
jacobi <param
echo 2000 >param
jacobi <param
echo 3000 >param
jacobi <param
echo 4000 >param
jacobi <param
echo 5000 >param
jacobi <param
echo 6000 >param
jacobi <param
echo 7000 >param
jacobi <param
echo 8000 >param
jacobi <param
echo 9000 >param
jacobi <param
echo 10000 >param
jacobi <param

Also, if you are having problems looking for good optimization options, 
I recommend looking at the gcc man page and see what compiler options 
get turned on with -O1, -O2 and -O3 (the man pages tell you explicitly 
what gets turned on with each level).


-- 
*Andrew J. Pounds, Ph.D.*
/Professor of Chemistry and Computer Science/
/Director of the Computational Science Program/
/Mercer University, Macon, GA, 31207 (478) 301-5627 /
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://theochem.mercer.edu/pipermail/csc435/attachments/20240301/f0368efa/attachment.html>