DSC-2025-23 | Hackathon: Explore Working on a (the DSC) HPC Cluster
Wann?
11. November 2025
13:00 - 17:30 Uhr
Wo?
Campus (Raum folgt in Kürze)
Trainer*innen
Nele Fuchs
Annika Nolte
Nils Leusmann
Timo Laudi
Data Science Center, Universit?t Bremen
Anzahl Teilnehmende: Max. 25
Sprache: Englisch
Background
High Performance Computing (HPC) clusters make it possible to tackle tasks that go far beyond the capabilities of a single laptop or workstation. Whether it’s processing massive datasets, training advanced machine learning models, running complex simulations, or performing large-scale data analysis, a cluster provides the speed, capacity, and parallel processing power needed to get results faster and more efficiently. Learning to work on an HPC cluster opens doors to solving problems at a scale that’s simply not possible on standard hardware.
The DSC HPC cluster consists of two nodes. Each node has two (3.0GHz) AMD EPYC 7313 16-Core Processor and 128GB RAM. In addition to that the specialized GPU Node is equipped with three Tesla V100S graphic cards, perfect for huge amounts of data used in data science or AI.
Event Format
This is a hackathon-style event. We’ll start with an introduction covering:
- HPC basics and why clusters are powerful tools
- The DSC cluster’s architecture and available resources
- How to connect, transfer data, and submit jobs
After that, you take the lead:
- Work with our cluster manual, on your own ideas or bring an existing project
- Team up with other participants to collaborate
- Explore provided example use cases such as BERT for large-scale text analysis, audio transcription with Whisper, and more examples showcasing how to work on the cluster.
- Get on-demand support from the DSC team for technical questions
The hackathon has a flexible ending. Participants are welcome to stay longer, keep working, wrap up at their own pace, and order some pizza.
Target Audience & Prior Knowledge
This hackathon is open to anyone curious about high performance computing and working on the HPC of the DSC. No prior HPC experience is required. The introduction will provide the basics, and the hackathon format allows everyone to work at their own pace and skill level
Technical Requirements
- Please bring your own notebook with a connection to the Wifi (e.g. via eduroam: /en/zfn/wifi/overview-wifi).
- A ZfN University Bremen Account (otherwise you can’t access the cluster).
- Optional: a project or problem to work on – otherwise, interesting example cases will be provided.
About the Trainers
Annika Nolte and Nele Fuchs are data scientists for training and consulting at the DSC. Nils Leusmann and Timo Laudi work as research associates at the DSC.
Nele Fuchs studied Philosophy, Material Culture: Textile (CvO University of Oldenburg), and Transcultural Studies (University of Bremen). As a data scientist in the Humanities, she supports researchers in the areas of Digital Humanities, data science methods for qualitative research and FAIR-compliant qualitative data management, leveraging her expertise in handling sensitive qualitative data.
As a DSC data scientist and environmental scientist, Annika Nolte supports researchers with their data management and analysis workflows. In training and consulting, Annika draws on broad expertise in Earth system sciences and extensive experience in scientific programming. Her main focus areas are data standardization, data management, statistical methods, geospatial analysis, and machine learning in environmental and marine sciences.
Currently pursuing a Ph.D. at the DSC, Nils Leusmann focuses on self-assessment routines for online algorithms. He holds a Master’s degree in Computer Science from the University of Bremen (2020) and a Bachelor's degree in Information Technology from DHBW Stuttgart (2016). His expertise includes digital twins, robotics, cognitive architectures, and knowledge representation. Nils has also worked extensively with Git for version control of his own research and collaborative development in scientific computing.
As a computer scientist, Timo Laudi manages the infrastructure and IT services of the DSC and offers support to their users. He has a particular interest in machine learning and, as part of his doctoral research, focuses on techniques and methods aimed at making artificial intelligence more sustainable and efficient.