Slurm Partitions (Compute Resources)

In Slurm, jobs are submitted to partitions – think of these as job queues for different types of jobs.

In your jobscript, you must specify the partition, number of cores and max “wallclock” time your job is allowed to run for (up to a permitted maximum time.) There is now no default wallclock time – you must specify it.

You may also need to specify other flags depending on which parition you submit to. For example, if using one of the GPU partitions, you’ll need to say how many GPUs you want. The HPC Pool and H200 partitions also require an account code.

The table below provides an overview of the partitions, and the compute resources, currently available in the upgraded CSF3 cluster (Slurm).

The Slurm jobscript template (shown further down this page) gives you a starting point for your jobscript. Much more detail on Slurm jobscript options is available in the sbatch command page and, for users of previous CSF systems, the SGE to Slurm reference.

You will also need to use Slurm commands on the login node: sbatch, squeue, scancel and srun.

-N numnodes
-n numtasks

-c corespertask

-A hpc-proj-code
-N numnodes
-n numtasks

-c corespertask

*Information is correct as of 1st October 2026

Jobscript Template

All jobs must specify the partition name (see above) and wallclock runtime limit (see the various time formats) for the job.

Note that the Slurm lines begin with #SBATCH – not #$ as used on SGE. A common mistake is to use a $ symbol in the Slurm special line. It is NOT #$BATCH.

Single-node CPU Jobs

#!/bin/bash --login
#SBATCH -p partitionname    # Required (all jobs) - see table above
#SBATCH -n numcores         # Required (parallel jobs)  - defaults to 1 for serial
#SBATCH -t timelimit        # Required (all jobs) wallclock timelimit. Max permitted is 7 days (7-0)

#SBATCH flag resources      # See table above for additional high-memory, CPU architecture
                            # and GPU resource flags.

# We recommend purging your env
module purge
module load apps/gcc/someapp/1.2.3

# $SLURM_NTASKS will be set to the number of cores requested above, if your app wants to know.
someapp.exe

# Or, if running an MPI app, Slurm knows to run $SLURM_NTASKS MPI processes if using mpirun
mpirun some_mpi_app.exe

For OpenMP parallel applications you can use:

#!/bin/bash --login
#SBATCH -n 1           # 1 task ($SLURM_NTASKS set to 1)
#SBATCH -c numcores    # cores-per-task ($SLURM_CPUS_PER_TASK set to this)
#SBATCH -t timelimit   # Max wallclock time

# Inform OpenMP how many cores to use
export OMP_NUM_THREADS=$SLURM_CPUS_PER_TASK
some_omp_app.exe

Multi-node CPU Jobs

Please note: the multinode partition is NOT available to free at point of use users (those in fatpou01) due to core limits applied to those users.

You must specify the -N numnodes and -n numtasks flags. You may also add the -c corespertask flag.

For a MPI-only application (no OpenMP capability), where you run an MPI process on every core of each node assigned to your job:

#!/bin/bash --login
#SBATCH -N numnodes    # Number of compute nodes (minimum is 2)
#SBATCH -n numtasks    # Number of MPI processes ($SLURM_NTASKS set to this)
#SBATCH -t timelimit   # Max wallclock time

module purge
module load apps/gcc/someapp/1.2.3

# Slurm knows to run $SLURM_NTASKS MPI processes if using mpirun.
mpirun some_mpi_app.exe

For a “mixed-mode” MPI+OpenMP capable application, where you run a small number of MPI processes but each uses OpenMP to do multicore computation on the compute node:

#!/bin/bash --login
#SBATCH -N numnodes    # Number of compute nodes (minimum is 2)
#SBATCH -n numtasks    # Number of MPI processes ($SLURM_NTASKS set to this)
#SBATCH -c numcores    # cores-per-task ($SLURM_CPUS_PER_TASK set to this)
#SBATCH -t timelimit   # Max wallclock time

module purge
module load apps/gcc/someapp/1.2.3

# Say how many cores each MPI process should use for OpenMP processing
export OMP_NUM_THREADS=$SLURM_CPUS_PER_TASK

# Slurm knows to run $SLURM_NTASKS MPI processes if using mpirun.
mpirun mpi-mapping-flags some_mpi_app.exe
          #
          # For one MPI process per node, use:
          # --map-by ppr:1:node --bind-to none
          #
          # For two MPI process per node (one per socket), use:
          # --map-by ppr:1:socket

GPU Jobs

Please see the GPU Jobs page.

High-memory Jobs

Please see the High Memory Jobs page.

Submitting Jobs and Checking your Queue

Submit the job using

sbatch jobscript

Check on the status of the job using

squeue

More jobscript options

For much more information on Slurm jobscript options, compared to SGE options, please see the SGE to Slurm reference.