The field of High Performance Computing (HPC) aims to solve scientific problems using high-performance computers (clusters), which far exceed the computational capacities (such as CPU performance, main memory and hard disk space) of normal desktop PCs or workstations.

The architecture of these computing systems often requires specific programming and software optimization. Individual calculations can last from a few hours to several weeks or months.

Most current HPC systems run the Linux operating system and they are operated using a shell via a command-line interface (CLI). Job scripts, which are text files containing information about the required computing resources, are used to submit individual calculations. These scripts are passed to a Job Scheduler, which then plans and coordinates the execution of the calculations.

Furthermore, users can work interactively by logging on to a compute node of the cluster directly. This provides the possibility to work with applications that have a graphical user interface (GUI) or to run other commands directly on a compute node.

This depends on the program in question and unfortunately there is no general answer. In principle, there is a good chance that your program will run on an HPC cluster if it runs on Linux, can be operated via the command-line interface, and it can compute in parallel on several CPU cores (keywords: OpenMP, MPI). This does not normally apply to commercial software that is only available for Windows and that can only be operated via a graphical user interface (GUI).

Researchers should be confident using a Linux command-line interface. Many universities have special introductory courses on how to use their local clusters. In addition, a detailed video tutorial introducing Linux in the context of HPC is available via HPC.NRW.

The use of the HPC systems is free of charge.

Licenses for commercial software must be obtained by the workgroups themselves if cluster- or campus-wide licenses are not available on-site. Additionally, a selection of software funded by the HPC.NRW initiative is available at all participating universities. More information can be found here.

Access to the systems is regulated by the individual sites and is only possible for registered users, usually via private SSH keys. Some universities also only allow access from their own university network and also block outgoing connections to the Internet.

Data copied to a cluster is usually only accessible to the owner and the administration team. However, it is notencrypted. If you need to process personal or sensitive data, you should ask your institution how to handle it beforehand.

The access criteria vary depending on the site providing the resources. At larger Tier-2 centers, computing time is usually allocated per project after evaluation by a standardized review process. Applications for such projects can generally be submitted several times a year. At smaller Tier-3 centers, access to the resources is often more straightforward but usually only available to working groups affiliated with the local university.

Detailed information about access requirements can be found in the Quick Reference Cards of the respective universities.

Yes, members of German public or state-funded teaching and research institutions can apply for computing time at Tier 2 centers in North Rhine-Westphalia. Depending on the type of application, a technical and/or scientific review will be carried out before the resources are granted. You can find more information about applying for computing time here. 

There are three Tier 2 centers located in NRW: RWTH Aachen University, Universität Paderborn und University of Cologne.

No, unfortunately only members of German public or state-funded teaching and research institutions can apply for computing time at German Tier 2 or Tier 3 centers. However, an exception may be made under certain circumstances if you collaborate with such an institution on a research project. We would be happy to advise you on this topic, please contact our helpdesk. 

Yes, in general members of German public or state-funded teaching and research institutions can apply for computing time at all Tier 2 centers in Germany. Depending on the type of application, a technical and/or scientific review will be carried out before the resources are granted. The Tier 2 centers of HPC.NRW are located at RWTH Aachen University, Universität Paderborn und University of Cologne. You can find information on how to apply here. 

Yes, you can contact the German Tier 1 centers or the European Tier 0 centers for computing time. The Gauss Center for Supercomputing is responsible for the allocation of computing time in Germany and PRACE organizes applications intended for European centers.

If you already use resources provided by an HPC center belonging to the HPC.NRW competence network, their local datacenter remains your initial point of contact as they can provide you with site-specific assistance. You can find the relevant contacts here. If the site with which you are affiliated does not have its own HPC resources (e.g. employees from universities of applied sciences), you are welcome to contact our helpdesk directly by email.

This depends on the chosen center. In HPC.NRW, the current regulations are as follows:

• RWTH Aachen University 
• University Paderborn
• University of Cologne 

All applications to Tier 2 centers should be submitted via the JARDS online application system. You can find more information about applying for Tier 2 computing time here:

• RWTH Aachen University 
• University Paderborn
• University of Cologne 

Please refer to our scalability tutorial on the HPC Wiki, which will provide you with all of the necessary information.