Info

We are excited to announce the continuation of the successful collaboration between HPC.nrw and MathWorks for the fourth consecutive year, offering a comprehensive series of four workshops designed to empower researchers, engineers, and scientists. This unique series aims to enhance skills in high-performance computing (HPC) and software development for research with MATLAB and SIMULINK.

This session Using  MATLAB on HPC Systems is the second part of the Matlab Workshop Series focuses on walking users through accessing the cluster, setting up a cluster profile, and submitting sample jobs. MATLAB is launched on the local workstation and parallel jobs are submitted to the cluster interactively, where users are shown how to set job submission arguments before running both single node and multi node jobs. Users will follow along and submit various CPU-only jobs before discussion moves to using GPUs. Additionally, users will learn how to debug and troubleshoot if any errors appear. 

Workshop Series

    1. Parallel Computing with MATLAB
    2. Using MATLAB on HPC Systems (this course)
    3. Introduction to Research Software Development with MATLAB
    4. Introduction to Research Software Development with SIMULINK

Organization

  • There is no seminar fee.
  • Presentations will be given in English. Slides will be available after the event.
  • This is an online event and will be held in Webex. Links are sent to registered attendees after registration.
  • You can/must register per topic, i.e., for each workshop separately.

Speaker

  • Damian Pietrus (Mathworks)

Course level

  • Intermediate

Target audience

  • Students, researchers and scientists with increasing computational demands

Prerequisites

  • Intermediate MATLAB familiarity (topics of "Parallel Programming with MATLAB")
  • MATLAB Client and Parallel Toolbox installed on your local Workstation/Laptop
  • For testing the hands-on demonstrations yourself:
    • Access to one of the clusters of Ruhr-University Bochum, RWTH Aachen University, Technical University of Dortmund, or University of Duisburg-Essen
    • Potentially VPN access (or other prerequisites) for the respective cluster)

Gained skills

  • How to configure MATLAB to submit jobs to a cluster 
  • The MATLAB job submission workflow 
  • Job optimization and tuning 
  • Using GPUs with MATLAB  
  • Troubleshooting job submission techniques 
  • Best practices for rehosting code onto the cluster

Course Material

  • will be published close to the event