Application for BayernKI access at NHR@FAU

Together with LRZ, NHR@FAU is providing Bavarian universities with easy access to basic compute resources (CPU & GPU) within the framework of “BayernKI”. For extended demand, think about submitting an NHR application at NHR@FAU or any of the other NHR centers. Demand going even beyond that may be served by the Gauss Center for Supercomputing.

Both LRZ and NHR@FAU have ordered dedicated hardware for BayernKI which is expected to arrive in late 2024 or early 2025. In the meantime, BayernKI users at NHR@FAU have limited access to few existing resources of “Alex” and “Fritz”. Groups from FAU or with a granted NHR project are not eligible for BayernKI at NHR@FAU, until the dedicated hardware is operational. The current phase of BayernKI at NHR@FAU is to make new groups familiar with the environment. Thus, stay tuned until early 2025 for real production and broader access within BayernKI.

For the time being, access within BayernKI will be granted on the level of organizational units (chairs, professorships, labs, etc.) and not project-based. You have to fill in the form below. The principal investigator (PI) and also the primary point of contact (PoC) have to log into NHR@FAU’s HPC portal before we can activate the project. If you are the first ones from your university, your local IT center may have to become active to allow the transfer of the required SSO attributes. Just follow the instructions in the HPC portal.

University, e.g. "TH-Deggendorf"

Organizational Unit (Chair/Lab/Professorship), e.g. "Professur für Robotik"

Webpage of the Organizational Unit

Name and title of the principal investigator (PI) (is responsible for compliance with the export control laws and, thus, should be supervisor of the people using the systems)

Email of the PI

optional: Name of the primary point of contact for technically handling the project (PoC) (Management of accounts; our point of contact for escalations)

optional: Email of the PoC

PI (and PoC) logged into NHR@FAU's HPC portal and accepted the terms of use and the export control declaration
yes

Scientific domain / research area based on DFG-Fachsystematik der Jahre 2016-2019, e.g. 409-05 for Interactive and Intelligent Systems, Image and Language Processing, Computer Graphics and Visualisation

Compute requirements, e.g. 8.000 A100 GPU-h within 8 months; typically 2 GPUs per job

Storage requirements, e.g. 4 TB of HDD storage for preprocessed data in webdataset tar files

Software requirements, e.g. Tensorflow 2.14

Very brief description of the planned work

Do you think you will require professional assistance for porting and optimizing your application on the target system?
yesno

Do you plan to process data which requires special protection (e.g. because of "patient data" or cooperation contracts - if you tick yes, further discussions will be required before the project can be activated. The current resources of LRZ and NHR@FAU are not really prepared to process such data.)
yesno

Other remarks

Please enter the answer for
1+5

Once your application is processed, you’ll get an email with further information. They will typically take a few working days.

For acknowledging the resource utilization, please follow the text for BayernKI on https://doc.nhr.fau.de/acknowledgment/.

NHR@FAU will contact you typically yearly for a report on your resource usage since the last report and the scientific outcomes of the resource utilization.