EEC High Performance Computing

To run Abaqus on a single compute node you can use the following procedure:

Login to HPC via terminal (e.g. use PuTTY or anything else)


module purge
module load abaqus

slurm submission script for abaqus:

#!/bin/bash
#SBATCH --time=24:00:00
#SBATCH --nodes=1
#SBATCH --partition=SMP
#SBATCH --ntasks-per-node=16
#SBATCH --job-name="mysimulation"
#
# input filename without extension
INPUT="input_filename_without_extension"

module load abaqus

EXECUTABLE=abaqus

#create scraqtch folder on all target nodes:
srun mkdir -p /scratch/${SLURM_JOBID}

##some parameters if needed for executable (delete the ones not needed)

ARGS="cpus=${SLURM_NTASKS} scratch=/scratch/${SLURM_JOBID} -verbose 1 user=userfile.f standard_parallel=all mp_mode=threads interactive"

# launch it
${EXECUTABLE} job=${INPUT} input=$INPUT.inp ${ARGS}
sleep 10

while [ -f ${INPUT}.lck ]; do
  sleep 5
done

# terminate gracefully
abaqus terminate ${INPUT}

# clean up
rm -fr /scratch/${SLURM_JOBID}

save this slurm script file e.g. as “abaqus.slurm” in your working folder containing all your input files and submit with
sbatch abaqus.slurm