HPC-Supercomputer „Noctua“ at NHR-site Paderborn

High-Performance Computing Centres

Support in applying for computing time and migrating your projects to national high-performance data centres
HPC-Supercomputer „Noctua“ at NHR-site Paderborn
Image: Universität Paderborn, Thorsten Hennig

Performance hierarchy of HPC centres

Tiering levels of high performance computing centers

Image: Yahyapour/Uni Göttingen

The hierarchy of European HPC infrastructure is typically divided into four tiers, with Tier 0 at the top and comprising European flagship systems such as the EuroHPC supercomputers.

  • Tier 1 (national level) represents the primary high-performance resource for research within a country and serves as a bridge between the European top tier and regional clusters. This tier includes the HPC clusters of the computing centres LRZ, HLRS and Jülich, which form the Gauss Centre for Supercomputing de.
  • Tier 2 is defined by regional competence centres or state computing centres that provide specialised architectures for local research projects and the training of early career researchers. This tier includes the HPC clusters of the NHR Centres de.
  • Tier 3 forms the foundation of the performance hierarchy and includes the university's internal IT supply as well as the institute's own clusters, which are essential for daily business and software development. Our HPC clusters Draco de and Ara de belong to this performance class.

NHR high performance computing centres (Tier 2)

Logo of NHR association

Graphic: NHR-Verein e.V.

NHR stands for National High Performance Computing. This is an association of computing centres at German universities that pool their resources and expertise to provide scientists throughout German with access to supercomputers.

There are currently nine NHR Centres spread across Germany. The Centres work together in the NHR Association. The aim is not for everyone to do everything, but for the centres to specialise. Each centre therefore usually has a specific area of expertise.

Geographic Distribution of NHR Centers

Geographic Distribution of NHR Centers

Image: https://nhrsw.de

The locations and their focus:

  • Aachen (RWTH): Fluid Mechanics, Chemistry
  • Berlin (ZIB): Renewable energies, Mathematics, Data analysis
  • Darmstadt (TU): Engineering sciences, material sciences
  • Dresden (TU): Life Sciences, Earth System Research
  • Erlangen-Nuremberg (FAU): Atomistic simulations, performance optimisation
  • Göttingen (GWDG): Humanities, Bioinformatics
  • Karlsruhe (KIT): Particle physics, engineering
  • Munich (LRZ/TUM): Astrophysics, Geophysics, AI
  • Paderborn (UPB): quantum computing, photonics

Applying for computing time

Researchers at German institutions of higher education can apply for computing time at the NHR Centres at any time.

We can support you in applying for computing time and in migrating your data and projects to and from these computing centres. Additional technical support is provided by the experts at the NHR Centres.

National supercomputers (Tier 1)

Logo of the Gauss Supercomputing association

Graphic: Gauss Supercomputing Verein e.V.

In Germany, three Centres form the national pinnacle of HPC infrastructure, united in the Gauss Centre for Supercomputing (GCS). These Centres represent Tier 1 and represent the highest level of performance for German science, while also acting as sites for European Tier 0 systems:

  • Jülich Supercomputing Centre (JSC): the JSC is home to JUPITER, Europe's first exascale supercomputer, which achieves a computing power of over one ExaFlop/s (1018 operations per second). The system has a modular design and was put into operation in 2025/2026 as one of the world's most powerful and energy-efficient systems.
  • High Performance Computing Centre Stuttgart (HLRS): HLRS inaugurated the Hunter system at the beginning of 2025, which serves as an interim solution and accelerator platform. It paves the way for the future exascale system Herder, which is planned for 2026/2027 and will be particularly optimised for complex simulations in the engineering sciences and AI applications.
  • Leibniz Supercomputing Centre (LRZ) in Garching: The LRZ currently operates the SuperMUC-NG (phases 1 and 2). While the first phase focussed on classic CPU performance, phase 2 increasingly integrates GPU accelerators to support modern AI workflows and quantum computing simulations within the national research landscape.

Advisory service and support

Researchers at German institutions of higher education can apply for computing time for the annually published Gauss Calls. The application is a competitive process. In particular, it must be demonstrated that such a large computing system is required for the project and that the application programmes have the necessary parallel scaling across many computing nodes.

We can support you in applying for computing time and in migrating your data and projects to and from these data centres. Additional technical support is provided by the experts at the Gauss Centres.

Contact us

HPC Consultancy

Support & Help
To Service-Desk
Link to download vCard
vCard