Access

A valid LRZ user account is needed to access the AI Systems.
An overview of LRZ user account options is given on the following page: Kennungen für LRZ-Dienste.
Master users are encouraged to reuse existing user IDs whenever possible.

Alternatively, access can be requested via the BayernKI initiative: https://www.ki-in-bayern.de/.

Prospective users of the AI Systems need to demonstrate the research background/focus of their intended work or use case.
This can be done by associating your user account with an AI-enabled LRZ project, i.e. a project with AI Systems user management capabilities.

The following outlines different scenarios: 

• If you know of an AI-enabled LRZ project that aligns with your research,
  contact the master user(s) of that project to provide your account access to the AI Systems.
  
  
• If there is a LRZ project that can be AI-enabled,
  the master user must provide additional information by submitting a service request to the LRZ Servicedesk.
  (choose "Predefined service request" → "AI Systems Project Activation/Update").
  Based on this information project suitability will be determined.
  
  
• If no suitable LRZ project exists,
  prospective master users can apply via Antrag auf ein LRZ-Projekt, 
  adding "AI Systems" under "Sonstige Bemerkungen" and provide the required information.
  After the project is created, submit a follow-up service request at the LRZ Servicedesk
  (choose "Predefined service request" → "AI Systems Project Activation/Update").
  Alternatively please refer to: https://www.ki-in-bayern.de/

Master users for AI-enabled LRZ Projects can manage individual accounts by adding or removing them from the <projectID>-ai-c group.
Membership in this group grants AI Systems access, use of general system resources, and VPN access to the Munich Scientific Network (MWN).
For more details, see How to add or remove members from groups?

Please bear in mind that due to the limited amount of currently available resources, the LRZ AI Systems should not primarily be considered as a development, prototyping or debugging environment. Any programs and scripts defining your production jobs should ideally be tested on other systems before. If you decide to use LRZ AI Systems resources for this purpose, choose adequately limited resources, preferably via https://login.ai.lrz.de (restricted to single GPUs/MIG devices; see 10. Interactive Web Servers on the LRZ AI Systems) or via suitable interactive allocation (see 6 Running Applications as Interactive Jobs on the LRZ AI Systems). Keep idle time at a minimum.

Regulations, Terms and Conditions

1. The usage regulations / Benutzungsrichtlinien 
   1. General regulations, see download center -> our regulations.
   2. The usage of LRZ infrastructure for processing of data, according to Art. 9 DSGVO "processing of special categories of personal data"

      The use of the LRZ infrastructure for the processing of data, according to Art. 9 DSGVO "processing of special categories of personal data", is generally not permitted. However, exceptions are possible if a special agreement is concluded with the LMU/TUM in addition to the existing data processing agreement ("Auftragsverarbeitungsvertrag", AVV), and on the other hand, the installation and operation of a specially secured infrastructure is usually required.

      In this case, please contact the LRZ via a service request (https://servicedesk.lrz.de/en) addressing the type of data you intend to process and requesting such a special agreement.

2. The usage of LRZ HPC systems is subject to export control regulations, see download center -> export control directives.
3. The LRZ publishes and stores your project data and description electronically. On explicit request this requirement may be waived.

Getting Started

In order to access the AI Systems, make sure you are connected to the Munich Scientific Network (MWN), then use SSH to contact login.ai.lrz.de. Provide your LRZ credentials to log in. 

For example, a user with an LRZ account xyyyyzz and a workstation connected to the Munich Scientific Network (MWN), can perform the login using the OpenSSH command line tools as follows:

ssh login.ai.lrz.de -l xxyyyzz 

This login node is meant for preparing and submitting your jobs to one or more of the available resources (see General Description and Resources), which are managed by the Slurm Workload Manager. For further details on this approach see Running Applications as Batch Jobs on the LRZ AI Systems.

Alternatively, a web-based frontend is accessible from within the Munich Scientific Network (MWN) at https://login.ai.lrz.de as well. The same credentials can be used for login. For futher details see Interactive Web Servers on the LRZ AI Systems.

Acknowledgement

Please use the following template when acknowledging the ressources and AI Systems-related support provided by LRZ: