Skip to content
Mordecai-Mark Mac Low edited this page Apr 22, 2024 · 7 revisions

How to run AMUSE on the Snellius supercomputer.

Snellius (https://servicedesk.surf.nl/wiki/display/WIKI/Snellius) is the Dutch national supercomputer.

Installing AMUSE on Snellius:

module load 2020 fosscuda/2020a
git clone https://github.com/amusecode/amuse.git
cd amuse
python3 -m venv env
source ./env/bin/activate
pip install --upgrade pip
pip install -r requirements.txt
pip install matplotlib astropy scipy

./configure --enable-cuda
make framework
make {codename}.code  # repeat for all codes you will be using
pip install -e .

Running an AMUSE script on Snellius:

Create a file (e.g. test.sh) with the contents below:

#!/bin/bash

# usage: sbatch test.sh

# SLURM settings
#SBATCH -J AMUSE_test
# Run for 3 minutes max
#SBATCH -t 00:03:00
# Request one node
#SBATCH -N 1
# Request 'thin' queue (or change to 'fat' or 'gpu')
#SBATCH -p thin
# #SBATCH -p fat
# #SBATCH -p gpu
# When using the 'gpu' queue, specify the number of gpus
# #SBATCH --gpus-per-node=1
# Write stdout/stderr to these files
#SBATCH --output=slurm_%j.out
#SBATCH --error=slurm_%j.out

# Load settings
module load 2020 fosscuda/2020a
source amuse/env/bin/activate

# run AMUSE script
python test_amuse.py

Make sure you have your test_amuse.py script in the same directory.