You gain access to an evolving range of systems hosted by our consortium partners. For JSC, you typically gain access to JURECA, but Discussions to grant access to further systems of JSC are ongoing. For HLRS, you gain access to the Vulcan Cluster and Hunter.

"Like our systems, the specifications of power and pperformance evolve as newer systems are brought online. For the most up-to-date information, consult the below links for our member institutes:
JSC: https://apps.fz-juelich.de/jsc/hps/jureca/configuration.html
HLRS: https://kb.hlrs.de/platforms/index.php/Hunter_(HPE)

Yes, our HPC systems offer a variety of node types and configurations tailored to different workloads—such as standard compute nodes, GPU-accelerated nodes, or large-memory nodes. More technical configuration details are available in each provider’s system documentation.
JSC: https://apps.fz-juelich.de/jsc/hps/jureca/configuration.html
HLRS: https://kb.hlrs.de/platforms/index.php/Vulcan and https://kb.hlrs.de/platforms/index.php/Hunter_(HPE)

Depending on the used system, you have full control over the kind of available hardware that you use when you submit your job to the batch system. In many cases, job schedulers and submission scripts allow configuration parameters such as memory size or GPU usage to be specified. Specific options depend on the system you’re using and are outlined in the technical documentation for that system.

Eine breite Palette an wissenschaftlicher Software, Compilern und Bibliotheken ist auf den verschiedenen HPC-Plattformen vorinstalliert und wird dort gewartet. Diese werden über Umgebungsmodule verwaltet. Jedes Zentrum führt auf seinem Dokumentationsportal eine Liste der verfügbaren Software.
JSC: https://www.fz-juelich.de/en/ias/jsc/services/user-support/using-systems/software
HLRS: https://kb.hlrs.de/platforms/index.php/Software_Development_Tools,Compilers%26_Libraries

Yes, user-level installations are supported and encouraged. The support teams at each center can assist you with installation guidance.

HPC systems are designed to support a broad range of applications, including many domain-specific frameworks. If your application relies on specific dependencies or tools, chances are they are already available or can be installed in user space. For confirmation and setup guidance, consult the system documentation or contact the support team at each institution.
JSC:
https://apps.fz-juelich.de/jsc/hps/jureca/configuration.html
HLRS:
https://kb.hlrs.de/platforms/index.php/Software_Development_Tools,Compilers%26_Libraries

No, our HPC systems run on Linux-based environments and do not natively support Windows applications.If you rely on Windows-specific software, please consult the support team to explore alternatives or compatibility options.

Each user is assigned a dedicated storage quota that depends on the system and usage type. Storage is typically organized into different tiers (e.g., fast scratch space and archival storage). Usage quotas, file retention policies, and backup availability vary—detailed information is available in the respective center's documentation.

Data transfers within the HPC environment are generally free of charge. However, transfers to external or cloud-based systems may be subject to policies or limitations. Refer to the system documentation for specifics.

Integration of external storage (e.g., institutional or cloud-based) is not possible. Please consult the support documentation or contact the helpdesk to explore options.

Available storage varies by system and is divided into different types, such as fast-access scratch space and long-term archival storage. Refer to the storage section in the documentation for current limits.
JSC:
https://apps.fz-juelich.de/jsc/hps/jureca/environment.html#available-file-systems
HLRS:
https://kb.hlrs.de/platforms/index.php/Storage_usage_policy

Data transfer to and from the HPC system is typically performed via secure command-line tools such as `scp` (secure copy) or `rsync` over SSH. Detailed instructions and examples can be found in the technical documentation of each center.

Access is usually provided via SSH (Secure Shell) to dedicated login or frontend nodes. You’ll need user credentials and possibly an SSH key to connect. Some systems also support multifactor authentication. Account requests and access permissions are managed through each center's registration process and user portal, as outlined in their documentation.

All major HPC centers implement secure access through SSH and enforce strong authentication procedures. Additionally, institutions like HLRS are certified under security standards such as ISO 27001 and TISAX Level 3. Details are available in the respective security documentation.

It depends. Some HPC centers adhere to recognized compliance standards. For example, HLRS is certified under ISO 27001 and TISAX Level 3, reflecting adherence to strict information security and data protection practices. Compliance details for each center can be found in their official documentation or upon request.

Yes, data transfers are conducted using secure protocols such as SSH, which encrypt both authentication credentials and data during transit.

In most cases, you will need to bring your own software licenses. At JSC, commercial users are responsible for providing the required licenses. HLRS does provide compiler licenses where applicable, but for most commercial software (ISV codes), users must either host the licenses themselves or rely on pay-per-use options from the software vendors. Because licensing conditions depend on the specific software package and may vary between centers, we recommend checking the documentation or contacting support at JSC and HLRS for clarification.

Yes, both centers provide technical support. At JSC, users receive the same support services as academic users, including help with setup, job submission, and troubleshooting. HLRS offers support primarily via their ticketing system, where users can submit questions or issues directly to the team. In both cases, support is provided by experienced staff who are familiar with the systems and typical usage scenarios.
Please note that the exact procedures and scope of support may vary between centers, so we recommend consulting their documentation or support pages for full details.
JSC: https://www.fz-juelich.de/en/ias/jsc/services/industry-relations
HLRS: https://www.hlrs.de/for-users/trouble-ticket-submission

Yes, both centers offer extensive user documentation to help new users get started. Since documentation is updated regularly and may differ by system, we recommend checking the respective centers' websites for the latest information.
JSC: https://apps.fz-juelich.de/jsc/hps/jureca/configuration.html
HLRS: https://kb.hlrs.de/platforms/index.php/Platforms

Currently, neither JSC nor HLRS provides a public community forum for peer-to-peer support. If such resources become available in the future, they will likely be announced on the respective websites.

Both centers follow a pay-per-use model. JSC typically issues bills on a quarterly basis and calculates charges based on actual usage. Similarly, HLRS uses quarterly billing, tracking system usage at the level of computing projects, which include individual user accounts.
Billing procedures may differ slightly between centers, so it is recommended to consult the specific billing policies provided by JSC and HLRS.

No, both JSC and HLRS aim to keep pricing transparent. There are no hidden fees or unexpected charges beyond what is outlined in the user agreement or service contract. Costs are clearly defined based on compute usage and any additional services explicitly requested.

At JSC, the specific payment method and invoicing process can be defined within the contractual agreement. HLRS typically accepts bank transfers for payment.
Details regarding invoicing, terms, and accepted methods should be confirmed directly with each center during the setup phase of your agreement.

The cost of compute time varies depending on the system in use and the terms of the commercial agreement. Both JSC and HLRS calculate prices based on the type of hardware and usage time.
You can refer to their pricing pages or contact the centers directly for the most current information.
JSC: https://www.fz-juelich.de/en/ias/jsc/systems/supercomputers/apply-for-computing-time

No, both JSC and HLRS use a pay-per-use model, meaning that billing is based on actual usage and typically handled after the fact.
Exact billing timelines and conditions may vary by center and contract.

At both JSC and HLRS, users can retrieve their data using standard secure methods such as scp or rsync. These tools allow for encrypted transfer of files from the HPC systems to your local environment.
Please consult your user documentation or support contact for exact instructions.

Yes, both centers expect that users will retrieve their data when they no longer require HPC resources. The exact procedures may vary between the centres. Be sure to review the offboarding or exit procedures in your specific agreement with each center.

Yes, both JSC and HLRS support the ability to scale resources to match changing workload requirements. Resources can be requested on demand depending on availability and system limits.

There are no hard restrictions stated in general; however, practical limits may be defined by the system configuration and fair usage policies at each center.

Yes, both centers use queue-based job scheduling systems. JSC uses SLURM to manage job submission and resource allocation. HLRS employs PBSPro for batch scheduling. These systems prioritize jobs based on resource requests, time limits, and system load.
For more details on job scheduling and queue behavior, consult the user documentation at each center.
JSC: https://apps.fz-juelich.de/jsc/hps/jureca/batchsystem.html
HLRS: https://kb.hlrs.de/platforms/index.php/Batch_System_PBSPro_(Hawk)

Priority options are not generally available, but exceptions may be possible on a case-by-case basis. HLRS may consider special arrangements for urgent jobs, while JSC does not currently offer a formal priority queue for commercial users.

Yes. HPC systems provide tools to check job status, queue position, and runtime details. For example, if your system uses SLURM, you can use commands like `squeue`, `scontrol`, or `sacct`. Each system may have slightly different monitoring tools—refer to the system's user documentation for specifics.

Jobs are submitted via batch scripts using workload managers such as SLURM. These scripts specify resources (e.g., cores, memory, run time) and execution commands. Users should connect via a login node and use the appropriate submission command (e.g., `sbatch`). Tutorials and examples are available in each HPC center's documentation.
JSC:
https://apps.fz-juelich.de/jsc/hps/jureca/batchsystem.html
HLRS:
https://kb.hlrs.de/platforms/index.php/Batch_System_PBSPro_(Hawk)

Wall-time (maximum runtime) restrictions vary by system but typically range up to 24 hours per job. Always check the current limits on the HPC site’s scheduler documentation.

Most HPC systems use high-speed interconnects like InfiniBand, optimized for fast node-to-node communication. Network speed is designed to support large-scale parallel computations and fast data movement. Specific bandwidth figures and configurations are detailed in system documentation.

Yes, access to the external internet is typically restricted for security and performance reasons. Login nodes may allow limited outbound access (e.g., for software downloads), but compute nodes are usually isolated. Firewalls and access controls are in place to protect both the user and system. Access and file transfer is only possible for registered IP addresses.