Computing Infrastructure
This page provides an overview of the different types of computing and storage infrastructure available to researchers at the Technical University of Munich. Whether you need high-performance computing resources, cloud-based solutions, or specialized AI systems, TUM and its partners offer a wide range of options to support your research.
At the end of this page, you will find a decision guide to help you choose the right solution based on your computing specifications (CPU, RAM, GPU) and research requirements.
Available Infrastructure
| Name | Type of Service | Access Requirement | Intended Use | Storage/Transfer Options | Link |
|---|---|---|---|---|---|
| LRZ Linux Cluster | HPC Cluster | TUM affiliation, application required | High-performance computing, parallel processing | LRZ DSS | LRZ Linux Cluster |
| LRZ Compute Cloud | Virtual Machines | TUM affiliation, project application | Flexible computing environments, web services | LRZ DSS, object storage | LRZ Compute Cloud |
| LRZ AI Systems | GPU Cluster | Application required, peer review | AI/ML training, deep learning | Specialized storage, LRZ DSS | LRZ AI Systems |
| LRZ Quantum Computing | Quantum Systems Cluster | Special application | Quantum algorithm research | LRZ DSS, Globus | LRZ Quantum |
| Terrabyte | HPC Cluster | TUM account, DFL employee | Earth observational data science | LRZ DSS, Globus | Terrabyte |
| EOSC | Virtual Machines, Jupyter Notebooks | European research affiliation | Flexible computing environments | Federated storage | EOSC Portal |
| Jupyter4NFDI | Jupyter Notebooks | Affiliation with German University | Interactive data analysis | Integrated storage | Jupyter4NFDI |
| NHR | National HPC | Doctoral researchers can apply | Large-scale simulations | High-throughput storage | NHR Alliance |
| EuroHPC | European HPC | Competitive application, calls open regularly | Extreme-scale computing | Parallel file systems | EuroHPC JU |
| AWS | Commercial Cloud | Credit card/budget | Scalable cloud computing | S3, EBS storage | AWS |
| Microsoft Azure | Commercial Cloud | Credit card/budget | Cloud services, integration with M365 | Blob storage, disk storage | Azure |
| Google Cloud | Commercial Cloud | Credit card/budget | ML/AI services, BigQuery | Cloud storage, persistent disks | Google Cloud |
| Local Workstation | Personal Computer | Direct access | Development, small-scale analysis | Local disks, network drives | N/A |
| WSL (Windows) | Linux on Windows | Windows 10/11 | Linux tools on Windows | Windows file system integration | WSL Docs |
TUM-Specific Resources
Leibniz Supercomputing Centre (LRZ)
The LRZ is TUM's primary partner for high-performance computing and provides several tiers of computing and storage solutions:
- Linux Cluster: General-purpose HPC for parallel computing tasks
- Compute Cloud: Virtualized infrastructure for flexible deployments
- AI Systems: State-of-the-art GPU clusters for artificial intelligence research
- Storage Solutions: Tiered storage from "hot data" to long-term archiving
Support: The LRZ offers comprehensive documentation, helpdesk support, and regular training courses. Overview: The LRZ wiki contains information on all services provided: LRZ Wiki, Overview of Computing options Consultation: Contact the LRZ Servicedesk for personalized advice on which service fits your needs.
NFDI (National Research Data Infrastructure)
Germany's National Research Data Infrastructure provides domain-specific services:
- Jupyter4NFDI: Interactive computing environments
- NFDI4Ing: Infrastructure for engineering sciences
- Domain-specific tools: Specialized services for different research fields
Major Research Instrumentation (Großgeräteanträge)
For research groups requiring dedicated hardware, TUM supports applications for major research instrumentation through DFG funding programs.
Decision Guide
How to Choose the Right Infrastructure?
Consider these key factors:
1. Computational Requirements
- CPU-intensive tasks (simulations, modeling): LRZ Linux Cluster, NHR
- GPU-intensive tasks (AI/ML, deep learning): LRZ AI Systems
- Memory-intensive tasks (large datasets in RAM): High-memory nodes at LRZ
- Interactive analysis: LRZ Compute Cloud, Jupyter4NFDI
2. Data Volume and Storage
- < 1 TB: Local workstation, LRZ Compute Cloud
- 1-10 TB: LRZ Linux Cluster storage
- > 10 TB: Terrabyte, specialized storage solutions
- Long-term archiving: Data Science Archive (LRZ), mediaTUM
3. Data Sensitivity
- Public data: Any service
- Sensitive/Personal data: LRZ services with data protection agreement
- Medical/Human data: Specialized secure environments, consult researchdata@tum.de
4. Budget Considerations
- Free (for TUM researchers): LRZ services, NFDI services
- Application-based: NHR, EuroHPC (free but competitive)
- Commercial: AWS, Azure, Google Cloud (pay-as-you-go)
5. Expertise Level
- Beginners: Jupyter4NFDI, LRZ Compute Cloud (with GUI)
- Intermediate: LRZ Linux Cluster (batch systems)
- Advanced: Custom setups, cloud infrastructure
Getting Started
- Assess your needs: Determine your computational, storage, and data protection requirements
- Consult documentation: Review the linked resources for each service
- Request access: Follow the application procedures for your chosen service
- Get training: Attend LRZ courses or contact the TUM Research Data Hub for guidance
Support and Feedback
For questions about which infrastructure is right for your project:
- General RDM support: researchdata@tum.de
- LRZ Services: LRZ Servicedesk
- Domain-specific guidance: Check our Domain-Specific Knowledge sections
Further information: TUM Research Data Hub — Storage & Infrastructure
Note: Infrastructure availability and access procedures may change. Always check the official websites for the most current information.