Matlab runs very slow on Supercomputer

Hello,

I am trying to run a python code on a supercomputer that is updated by running a generalized eigenvalue problem that I wrote in Matlab at every time-step. For some reason, the matlab code is very slow, even slower than when run on my normal machine. I tried testing the matlab code individually by writing a shell script and it was still slow. The shell script to run the matlab code looks as such:

#!/bin/bash

#SBATCH --job-name=matlab_test

#SBATCH --mem-per-cpu=10

#SBATCH --ntasks-per-node=24

#SBATCH --cpus-per-task=1

#SBATCH --nodes=1

module load MATLAB/R2021a

matlab -nodisplay -nosplash -nodesktop -nojvm -r "run('my_matlab_code.m'); exit;"

I have noticed that the Matlab also takes a considerable amount of time to load. I was wondering if there is an issue with the configuration or the way I am calling Matlab on the supercomputer. Consequently, I think only one core is being used and I do not know why that is.

Thank you,

Jack

[Antoine]

Hello, I have the same issue, did you find where was the problem ?