NHR@Göttingen
Universität Göttingen
Göttingen

NHR@Göttingen - operated by GWDG together with University of Göttingen - provides a secure, modern HPC platform for data-intensive CPU and GPU applications with a special focus on usability. All services are supported by advice and training by domain experts.
Key Expertise and Profile
- Expert support and supplementary services especially for life-sciences and humanities, quantum computing, and AI applications
- Usability is enhanced by modern access methods like JupyterLab with multiple kernel options (Python, R, Julia, SageMath, etc)
- Remote desktops with 3D acceleration and online IDEs with AI support enable rapid prototyping and analysis – without the need for local installation
- Powerful GPU nodes and support from experts enable to train and test AI models, such as large language models (LLMs)
- SecureHPC is a data protection-compliant environment for sensitive data that meets the highest security standards – ideal for medical, social science or personal data.
Technical Specifications
| Center Name | NHR@Göttingen |
|---|---|
| Total Resources: | |
| # CPU-Cores | 136 320 |
| # GPUs | 168 |
| Additional Information | |
| max. job run time [h] | 168 |
| jupyter notebook available [y/n] | y |
| CPU Resources: | |
| System 1 | Emmy P3 |
| CPU-Type | Intel Sapphire Rapids 8468 |
| # nodes | 398 |
| # Cores per node | 96 |
| memory per node [GB] | 256 - 2048 |
| System 2 | Emmy P2 |
| CPU-Type | Intel Cascade Lake AP 9242 |
| # nodes | 1 022 |
| # Cores per node | 96 |
| memory per node [GB] | 384 - 1537 |
| GPU Resources | |
| System 1 | Grete P3 |
| GPUs per node | 4 x Nvidia H100 |
| # nodes | 5 |
| memory per GPU [GB] | 94 |
| System 2 | Grete P2 |
| GPUs per node | 4 x Nvidia A100 |
| # nodes | 34 |
| memory per GPU [GB] | 40 |
| System 3 | Grete P1 |
| GPUs per node | 4 x Nvidia V100 |
| # nodes | 3 |
| memory per GPU [GB] | 32 |