Research Blog
Welcome to my Research Blog.
This is mostly meant to document what I am working on for myself, and to communicate with my colleagues. It is likely filled with errors!
This project is maintained by ndrakos
Here is a summary of the 2048 simulations. I had previously documented my workflow here, but there were a couple of updates.
Here was the pbs job script I ended up using:
#PBS -l select=128:ncpus=8:model=has
#PBS -l walltime=120:00:00
#PBS -q long
%module load mpi-sgi/mpt
%module load comp-intel/2018.3.222
cd $PBS_O_WORKDIR
module load pkgsrc/2018Q3 mpi-sgi/mpt
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/nasa/pkgsrc/sles12/2016Q4/lib
cd /u/ndrakos/Gadget2/
mpiexec -np 1024 dplace -c CS ./Gadget2 /nobackup/ndrakos/wfirst2048/wfirst2048_gadget.param > /nobackup/ndrakos/wfirst2048_2/output
Here is a comparison of the 2048 simulation to the 512 simulation (the colour scheme isn’t ideal, but it shows they agree).
512:
2048:
To get a second realization, I changed all the seeds as follows (in the MUSIC .conf file):
[random]
seed[7] = 11111
seed[8] = 11111
seed[9] = 11111
seed[10] = 11111
seed[11] = 11111
seed[12] = 11111