This directory contains scripts for setting up, running and postprocessing the deal.II CEED BPs written by Martin Kronbichler.
Running the scripts
../../go.sh -c linux -m gcc -r bp1.sh -n 16 -p 16
-n 16 is the total number of processors and
-p 16 is the number of
processors per node.
Multiple processor configuration can be run with:
../../go.sh -c linux -m gcc -r bp1.sh -n "16 32 64" -p "16 32 64"
Post-processing the results
First, save the output of the run to a file:
../../go.sh -c linux -m gcc -r bp1.sh -n 16 -p 16 > run_001.txt
and then use one of the
postprocess-plot-*.py scripts (which require
the python package matplotlib), e.g.:
python postprocess-plot-2.py run_001.txt
Note that the
postprocess-*.py scripts can read multiple files at a time just
by listing them on the command line and also read the standard input if no files
were specified on the command line.
Plotting performance comparisons
postprocess-plot-4.py can be used to plot the performance ratios
between two sets of data, e.g. vector (BP2) vs scalar (BP1), or two compilers:
XLC vs GCC. Sample usage:
python postprocess-plot-4.py bp1_gcc_n16.txt bp2_gcc_n16.txt python postprocess-plot-4.py bp1_gcc_n16.txt bp1_xlc_n16.txt