Gary has achieved near perfect scaling efficiency:
https://mail.{{gna.org/public/mailing list url|relax-devel/2007-05/msg00000.html}}
=== Dependencies ===
# Show what loading does
module show openmpi-x86_64
# or
module show openmpi-1.10-x86_64
# See if anything is loaded
# Load
module load openmpi-x86_64
# Or
module load openmpi-1.10-x86_64
# See list
module list
# For 64 bit computer.
sudo ln -s /usr/lib64/openmpi/bin/mpicc /usr/bin/mpicc
# or
sudo ln -s /usr/lib64/openmpi-1.10/bin/mpicc /usr/bin/mpicc
</source>
Remember to check, if there are newer versions of [https://bitbucket.org/mpi4py/mpi4py/downloads/ mpi4py]. <br>
The [https://bitbucket.org/mpi4py/mpi4py mpi4py] library can be installed on all UNIX systems by typing:
<{{#tag:source lang="bash">|
# Change to bash, if in tcsh shell
#bash
v=2.0.0{{current version mpi4py}}
#tcsh
set v=2.0.0{{current version mpi4py}}
pip install https://bitbucket.org/mpi4py/mpi4py/downloads/mpi4py-$v.tar.gz
cd ..
rm -rf mpi4py-$v
</source>|lang="bash"}}
Then test
You can continue try this, until a good results
<source lang="bash">
# With shellmpirun -np 4 echo "hello world" # With pythonmpirun -np 4 python -m mpi4py helloworld # If newer version of mpirun, then --report-bindings worksmpirun --report-bindings -np 11 4 echo "hello world"
mpirun --report-bindings -np 12 echo "hello world"
# This is too much
# Set number of available CPUs.
set NPROC=`nproc`
set NP=`echo $NPROC + 1 0 | bc `
echo "Running relax with NP=$NP in multi-processor mode"
#set NPROC=`nproc`
set NPROC=10
set NP=`echo $NPROC + 1 0 | bc `
# Run relax in multi processor mode.