• Zur Metanavigation
  • Zur Hauptnavigation
  • Zur Suche
  • Zum Inhalt
  • Zur Subnavigation
  • Zum Seitenfuß
  • University of Hamburg
  • For journalists
  • Staff Service Portal0
  • STiNE
https://www.kus.uni-hamburg.de/aktuelles.json?recentnews=true
  • Institutions and target groups
    • Einrichtungen
      • Faculties

        • Faculty of Law
        • Faculty of Business, Economics and Social Sciences
        • Faculty of Medicine
        • Faculty of Education
        • Faculty of Humanities
        • Faculty of Mathematics, Informatics and Natural Sciences
        • Faculty of Psychology and Human Movement Science
        • University of Hamburg Business School
      • Services

        • Study Information Net STiNE
        • Surfmail
        • Exchange
        • Staff Service Portal
        • Learning Platforms
        • Katalogplus – Catalogue of Hamburg Libraries
        • SharePoint
      • Student facilities and services

        • Campus Center
        • Career Center
        • PIASTA
        • AStA (student council)
        • University Sports
        • University Music
        • Family Office
        • University Language Center
        • General Languages
        • Office of Affairs for Students with Disabilities or Chronic Diseases
        • Departmental libraries
        • Healthy Campus
      • Further facilities and services

        • The Cluster of Excellence CLICCS
        • The Cluster of Excellence CUI
        • The Cluster of Excellence Quantum Universe
        • The Cluster of Excellence Understanding Written Artefacts
        • Hamburg Research Academy
        • Pier Plus
        • Hamburg Center for University Teaching and Learning (HUL)
        • Hub for Crossdisciplinary Learning (HCL)
        • CEN
        • Museums and collections
        • University Archives (in German)
        • Regional Computing Center (RRZ)
        • University Administration
        • Knowledge Exchange Agency
        • The Unikontor Shop (in German)
    • Target groups

      • Prospective students
      • Students
      • International prospective students
      • Young researchers
      • Researchers
      • Instructors
      • Staff
      • Prospective continuing education students
      • Alumni
      • Journalists
      • Refugees (in German only)
    Close
  • /en/sitemap.json
University of Hamburg - to research, to teach, to educate and form, to homepage

RRZ

Regional Data Center
(RRZ)
  • Support
    • RRZ-ServiceLine/STiNE-Line
    • RRZ-ServiceDesk
    • RRZ-ServicePortal
    • RRZ Feedback
    • RRZ Portfolio
    • Application Forms
    • Instructions/Glossary
    • FAQ
    • Policies
    • Support Forms
  • Services
  • ID & Authentication
    • Apply for an identifier
    • Forgotten/change password
    • Set up two-factor authentication
    • Managing identifier groups (group manager)
  • Innovation
    • Research & Education
    • Organisational Development
  • About Us
    • RRZ News
    • Mission statement
    • Map / Directions
    • Operating Hours / Opening Hours
    • RRZ Staff
    • Careers
    • Organizational Structure
    • Webinars
    • Equal Opportunity at the RRZ
    • History of the RRZ

Photo: Thomas Orgis

  1. UHH
  2. RRZ
  3. Services
  4. HPC
  5. Hummel-2
  6. quick start
  7. gpu-test

GPU test node (front3)

The GPU front-end node front3 is provided for

  • interactive GPU access for development and testing,
  • testing GPU batch jobs.

For interactive work front3 is the equivalent of the CPU front-end nodes front1 and front2.

In addition, small batch jobs can be run on front3 be employing --partition=gputest (the default time limit is 10 minutes). The idea is that functionality can be tested. Since all resources of the node are shared all test runs should use as little memory (RAM and VRAM) as possible.

Recall that

  • RAM allocation can be checked with the free, top and ps -l commands,
  • VRAM allocation can be checked with the nvidia-smi command.

Running batch jobs in the gputest partition

A batch job that is running in the gputest partition can use GPU but must not specify GPU resources (--gpus must not be set, see the example job below).

Differences to the gpu partition

  • --gpus (or equivalent) must not be specified.
  • /home is writable.
  • Resources including the GPU are shared with other batch- or interactive jobs, hence time measurements are not reliable.

Example job

line
no.
/sw/batch/examples/gputest-partition/gputest-job.sh
1
2
3
4
5
6
7
8
9
10
11
12
#!/bin/bash
#SBATCH --account=WorkingGroupName_gpu
#SBATCH --partition=gputest
#SBATCH --export=NONE

source /sw/batch/init.sh

module load cuda/12.5.1

$CUDA_HOME/extras/demo_suite/busGrind

exit
  • Hummel-2 (HPC cluster 2024)
    • getting access
    • quick start
      • GPU test node (front3)
      • migration from Hummel
      • hardware
      • working with data
      • software
      • running batch jobs
      • RRZ tools
      • acknowledgements
    • Basic skills
    • Courses / events
    • Museum
    • Contact

    Last update: 5 February 2026

    • Feedback
    • Legal notice
    • Privacy statement
    • Accessibility
    • Site map

    University of Hamburg

    Innovating and Cooperating for a Sustainable Future in a Digital Age

    • Follow us on Instagram.
    • Follow us on Facebook.
    • Connect with us on LinkedIn.
    • Follow us on Mastodon.
    • Follow us on Bluesky.
    • Subscribe to our channel on YouTube.
      • Studies
      • Prospective students
      • Continuing education
      • Research
      • Teaching at UHH
      • Doctoral studies
      • Job opportunities
      • Knowledge exchange
      • University
      • Excellence
      • Faculties
      • Alumni
      • Contact
      • Campus Navigator
      • UHHApp
      • Emergencies, safety and security
      • Antidiscrimination
    Go to family-friendly higher education institution audit
    Go to internationalisation re-audit
    Go to the Sustainability Audit page
    Go to Diversity Audit
    Go to System accreditation

    © 2026 University of Hamburg. All rights reserved