CPU Examples

Standard job using a single CPU

#!/bin/bash
#    
#SBATCH -p wn
#SBATCH --account=t3
#SBATCH --job-name=job-name
#SBATCH --mem=3000M                     # memory 3GB (per job)
#SBATCH --time 01:00:00
#SBATCH -o %x-%j.out    # replace default slurm-SLURM_JOB_ID.out; %x is a job-name (or script name when there is no job-name)
#SBATCH -e %x-%j.err    # replace default slurm-SLURM_JOB_ID.err 


echo HOME: $HOME 
echo USER: $USER 
echo SLURM_JOB_ID: $SLURM_JOB_ID
echo HOSTNAME: $HOSTNAME

# You should use a node's local /scratch space for creating files and
# intensive I/O. Never use /tmp, and make sure that you also
# set TMPDIR, so that your programs create temporary files there
JOB_SCRATCH=/scratch/$USER/$SLURM_JOB_ID
mkdir -p "$JOB_SCRATCH"
export TMPDIR="$JOB_SCRATCH"

#########################################################
# Please fill in your code here


#########################################################
# cleaning of temporary working dir after job is completed:
rm  -rf "$JOB_SCRATCH"

date 

Jobs using multiple CPUs

Example of the job header for running 8 parallel threads within a single task, each thread getting it's own logical CPU (hardware thread).

#!/bin/bash 
#SBATCH --job-name=smp_example         # Job name
#SBATCH -p wn
#SBATCH --account=t3
#SBATCH --ntasks=1                          # Run a single task
#SBATCH --cpus-per-task=8                   # Number of CPU cores per task; Example of  submitting  8-core parallel SMP job
#SBATCH --mem=3gb                           # memory request for whole job
## Alternatively, you could set instead a memory per CPU requirement
##   #SBATCH --mem-per-cpu=3072

#SBATCH --time=02:00:00                     # Time limit hrs:min:sec
#SBATCH --output=smp_%j.log                 # Standard output and error log

your_script.sh