By: Marianne Fuglsang Welling Farsinsen, AAU Kommunikation og Public Affairs
Foto: Kasper Patrick Bährentz 

Supercomputing – also known as HPC (High Performance Computing) – has long been available in research for the development of AI, such as language models, training neural networks and working with large data sets. Now the university's students will also have the opportunity to gain experience with the practical use of computing power through AI-LAB which can be used for all education-related activities.

"We are proud to be able to offer AAU students competence development in HPC and practical experience with AI and handling large data sets. Both research institutions and the business community are looking for graduates with specialist skills in AI, so AI-LAB is an important part of best equipping our students to contribute with their knowledge and skills after graduation," says Pro-rector Anne Marie Kanstrup.

Get off to a good start at AI-LAB

The university's teaching staff before the summer holiday had the opportunity to try out and experiment with AI-LAB. Some teaching staff already have a lot of experience with HPC resources and machine learning through their own research, and for this group it will be easy to include AI-LAB in their teaching. For other teachers, the entire HPC area is relatively new, and good help is available. IT Services' CLAAUDIA is responsible for the operation and support of AI-LAB and is making a special effort to help teachers and students get started:

"We offer to come out and help teaching staff onboard their students to AI-LAB - just submit a request via the service portal! Students can also use AI-LAB in their project work, and we have made really thorough guides so that even novices can follow along," says Frederik Petri Svenningsen, CLAAUDIA, ITS.

In addition to teachers being able to try out AI-LAB over the summer, a number of students have also been in the loop with a view to making guides easy to use for the students, for example. Like Mikkel Ørts Nielsen, who is studying Business Data Science:"For someone with minimal code and terminal experience, this guide is good because it’s easy to follow and has clear explanations, so I don't just copy/paste without knowing what is happening," says Mikkel Ørts Nielsen.

How to get started

If you are a member of the teaching staff and want to start using AI-LAB, you must follow these steps:

• Fill out the access form (it can take up to two business days before you get access; you will receive an email when access is granted).
• Visit the AI-LAB documentation and follow the "Getting started" guide and learn how to log in, transfer files, download applications and how to run jobs.
• Explore the different guides and feel free to leave your feedback in this form.

Any questions or need help getting started?

CLAAUDIA offers to help onboard students, so if teachers want a session with a data scientist where their students get help to get started, contact CLAAUDIA via the service portal.

Questions or other help on using AI-LAB can also be directed to CLAAUDIA through the service portal.

What do the teachers say?

"In the "Design and Development of Artificial Intelligence" (DAKI) programme, the students are already in full swing using the AI-LAB platform which enables them to try out the techniques they learn in the course "Deep Learning" in practice, as well as effectively train AI models as part of their current and future semester projects." Andreas Aakerberg, teaching staff, Deep Learning, 3rd semester,  DAKI

"We would like to use AI-LAB to show students in our Master's programme in Business Data Science how to build business-relevant systems and solutions with local language models – especially systems where data is not allowed to be sent out to the cloud. Fine-tuning and manipulation of generative AI models will certainly also be part of many semester projects here in Autumn 2024. So, it’s important to have access to powerful hardware that enables working with these new and exciting technologies." Roman Jurowetzki, teaching staff, Applied Data Science