Start Here
Use the information provided here to get started with using the HPC clusters at GWDG, which are arranged by topic in the sidebar and below.
A cluster is a group of powerful computers called “nodes” working together, that typically have a very fast network connection between them and have access to shared storage. You use it by writing job scripts that describe the amount of nodes, CPU cores, memory, and/or GPUs you need and hand it to the scheduler. The scheduler matches your requirements against the available hardware and starts your job once sufficient resources can be allocated. After your job completes, you will get the results and an overview of the used resources. More details about our cluster can be found on the cluster overview page. Details about getting an account and a brief introduction to working with a scheduler are also explained in this documentation. If you need more experience, feel free to check out our courses.
You are from a specific science domain, have not used HPC before or are switching from another cluster? We have special starting pages for you! Search the list below to find your science domain and check what we offer for you specifically.
- Getting An Account
- Cluster Overview
- Types of User Accounts
- NHR Application Process
- Using AI Services
- KISSKI Application Process
- Connecting (SSH)
- Using the cluster
- Accounting Core-h
- Science Domains
If you have questions or run into problems, you can create a support ticket by sending an email with your question or problem to the appropriate email address. Everything you need to know should be listed on our support page.