Skip to main content

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

NameType of ServiceAccess RequirementIntended UseStorage/Transfer OptionsLink
LRZ Linux ClusterHPC ClusterTUM affiliation, application requiredHigh-performance computing, parallel processingLRZ DSSLRZ Linux Cluster
LRZ Compute CloudVirtual MachinesTUM affiliation, project applicationFlexible computing environments, web servicesLRZ DSS, object storageLRZ Compute Cloud
LRZ AI SystemsGPU ClusterApplication required, peer reviewAI/ML training, deep learningSpecialized storage, LRZ DSSLRZ AI Systems
LRZ Quantum ComputingQuantum Systems ClusterSpecial applicationQuantum algorithm researchLRZ DSS, GlobusLRZ Quantum
TerrabyteHPC ClusterTUM account, DFL employeeEarth observational data scienceLRZ DSS, GlobusTerrabyte
EOSCVirtual Machines, Jupyter NotebooksEuropean research affiliationFlexible computing environmentsFederated storageEOSC Portal
Jupyter4NFDIJupyter NotebooksAffiliation with German UniversityInteractive data analysisIntegrated storageJupyter4NFDI
NHRNational HPCDoctoral researchers can applyLarge-scale simulationsHigh-throughput storageNHR Alliance
EuroHPCEuropean HPCCompetitive application, calls open regularlyExtreme-scale computingParallel file systemsEuroHPC JU
AWSCommercial CloudCredit card/budgetScalable cloud computingS3, EBS storageAWS
Microsoft AzureCommercial CloudCredit card/budgetCloud services, integration with M365Blob storage, disk storageAzure
Google CloudCommercial CloudCredit card/budgetML/AI services, BigQueryCloud storage, persistent disksGoogle Cloud
Local WorkstationPersonal ComputerDirect accessDevelopment, small-scale analysisLocal disks, network drivesN/A
WSL (Windows)Linux on WindowsWindows 10/11Linux tools on WindowsWindows file system integrationWSL 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

  1. Assess your needs: Determine your computational, storage, and data protection requirements
  2. Consult documentation: Review the linked resources for each service
  3. Request access: Follow the application procedures for your chosen service
  4. 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:

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.