Day 1 - Workshop Parallel Computing with MATLAB
on Wednesday, 27/05/2026, 09:00 am - 12:00 am + 02:00 pm - 05:00 pm
Speaker: Dr.-Ing. Stefan Kerber, MathWorks
Modern research increasingly benefits from making effective use of multi‑core processors, GPUs, and clusters. In this workshop we introduce MATLAB language constructs that support scalable and efficient computational workflows.
You will learn how to write or modify MATLAB code for parallel execution on desktop systems, GPUs, and local clusters. The focus is on practical programming motifs that build on existing MATLAB knowledge and can be directly applied in day‑to‑day research.
Topics:
- MATLAB Multi-Threading and Multi-Core
- Language constructs for synchronous and asynchronous parallel programming
- Communicating (intermediate) results during parallel code execution
- GPU Programming with MATLAB
- Introduction to batch execution of parallel MATLAB code on clusters and clouds
Prerequisites
A beginner-level familiarity with MATLAB is expected for this course. The course will be conducted in an online environment, so no installed MATLAB is required. If you want to work on your own machine, make sure you have MATLAB R2025b including the Parallel Computing Toolbox installed. Note: MATLAB only supports NVIDIA GPUs for GPU programming.
Who is this for?
The workshop is for students, scientists, and researchers at universities in Saxony (and beyond) that face growing computational demands or volumes of data. You do not need to have a project for the HPC clusters of TUBAF or TUD yet. A general interest in using parallel hardware more effectively in your MATLAB workflows is sufficient.
Speaker Bio
This workshop is presented by Dr. Thomas Künzel. Thomas holds a PhD in Zoology/Animal Physiology from RWTH Aachen University and led a research group in auditory neuroscience before he joined the MathWorks. Thomas focuses on scientific- and high-performance computing applications with MATLAB. He supports researchers throughout central Europe in effectively using local and remote parallel resources for their computational projects.