==== MATLAB GPU job ====

=== Execution of a MATLAB program on GPU ===

In this example we see how to run the [[https://it.mathworks.com/matlabcentral/fileexchange/34080-gpubench|GPUBench]] MATLAB program on GPU. Requires the MATLAB Parallel Computing Toolbox and a GPU with CUDA Compute Capability.

Script ''slurm-matlab-gpu-bench.sh'':

<code bash slurm-matlab-gpu-bench.sh>
#!/bin/bash
#SBATCH --job-name=GPUBench
#SBATCH --output=%x.o%j
#SBATCH --error=%x.e%j
#SBATCH --nodes=1
#SBATCH --ntasks-per-node=1
#SBATCH --gres=gpu:1
#SBATCH --partition=gpu
#SBATCH --qos=gpu
#SBATCH --mem=8G
#SBATCH --time=0-00:30:00
#SBATCH --account=<account>

module load matlab/R2017a

cd 'GPUBench_v1.10'

matlab -nodisplay -nosplash -r "T=gpuBench;quit"
</code>

Submitting ''slurm-matlab-gpu-bench.sh'':

<code>
sbatch slurm-matlab-gpu-bench.sh
</code>

=== Execution of two MATLAB programs on the same GPU ===

In this example we see how to run two instances of the [[https://it.mathworks.com/matlabcentral/fileexchange/34080-gpubench|GPUBench]] MATLAB program on the same GPU. Requires the MATLAB Parallel Computing Toolbox and a GPU with CUDA Compute Capability.

Script ''slurm-matlab-multi-gpu-bench.sh'':

<code bash slurm-matlab-multi-gpu-bench.sh>
#!/bin/sh
#SBATCH --job-name=GPUBench
#SBATCH --output=%x.o%j
#SBATCH --error=%x.e%j
#SBATCH --nodes=1
#SBATCH --ntasks-per-node=1
#SBATCH --gres=gpu:1
#SBATCH --partition=gpu
#SBATCH --qos=gpu
#SBATCH --mem=8G
#SBATCH --time=0-00:30:00
#SBATCH --account=<account>

module load matlab/R2017a

cd 'GPUBench_v1.10'

pids=""

for i in $(seq 0 1); do
  matlab -nodisplay -nosplash -r "T=gpuBench;quit" \
  1>"$SLURM_SUBMIT_DIR/${SLURM_JOB_NAME}.o${SLURM_JOB_ID}.${i}" \
  2>"$SLURM_SUBMIT_DIR/${SLURM_JOB_NAME}.e${SLURM_JOB_ID}.${i}" &
  pids="${pids:+$pids }$!"
done

echo "Waiting PID(s): $pids"

wait $pids
</code>

Submitting ''slurm-matlab-multi-gpu-bench.sh'':

<code>
sbatch slurm-matlab-multi-gpu-bench.sh
</code>

[[calcoloscientifico:userguide:matlab|go back one level]]