top of page

Groningen Could Become the Silicon Valley of the North

  • Writer: Daniela Ivančáková
    Daniela Ivančáková
  • Jan 30
  • 3 min read

A regional investment and development agency has recently announced a feasibility investigation concerning the establishment of an AI factory in Groningen. But what does that really mean?


Almost one-fourth of people in the Netherlands use AI to generate text, videos or images © Daniela Ivančáková
Almost one-fourth of people in the Netherlands use AI to generate text, videos or images © Daniela Ivančáková

After the Dutch Minister of Economic Affairs, Dirk Beljaarts announced that the Netherlands is interested in an AI factory, the House of Representatives approved a motion, emphasizing the environmental responsibility. NOM, an Investment and Development Agency for the Northern Netherlands, is currently engaged in discussions with organizations, companies and entrepreneurs as Groningen remains a promising option for the location.


The project stands on the European Commission’s program EuroHPC. The aim is to build factories for developing energy-efficient AI technology and through that decrease the dependence on foreign big tech companies like Google or Microsoft.


In 2024, 23 percent of the Dutch population utilized AI programs, as stated in Statistics Netherlands’s report. The size of the Artificial Intelligence market in the Netherlands is expected to reach 2.55 billion US dollars in 2025, according to data provided by Statista.



The topic of AI has been largely discussed over the recent years, however, due to its abstract nature and complexity it might be difficult to follow. To shed more light on the matter, The Glass Room talks to an assistant professor at the Faculty of Science and Engineering of the University of Groningen, Matthia Sabatelli, Ph.D.

 

AI factory in a nutshell

“AI technologies need to be trained on powerful computers, and so they need some very specific hardware, which needs to be gathered all in one spot,” explains Sabatelli.

To provide a better picture of an AI factory, it is a number of expensive and very powerful machines put in a common location, accessible for the training of AI systems. While the University of Groningen already has a supercomputer, Hábrók, Sabatelli clarifies that it is not equipped to train modern AI algorithms.

 

Why Groningen?

Sabatelli outlines two reasons when it comes to considering Groningen as a convenient location for the project. Because of Hábrók, there is a lot of experience when it comes to maintaining a supercomputer. He also says that the AI faculty at the University of Groningen's wealth of scientific knowledge can help in the building process.

“We do a lot of research, which uses the type of hardware, which will be put in this AI factory. So, it's the perfect scenario if you want to build something like that. At some point, you will also have to test it. And we have a lot of people who can do it,” he tells The Glass Room.


According to a press release by NOM, the presence of an AI factory can support software and AI companies, projects like UMCG’s Health Data Valley and also public data processors, such as DUO. Offering "good and green computing", the AI factory apparently has great potential to also attract new companies that value independence, sustainability and security of datasets.

 

Funding

While the European Commission will finance half of the costs, it is up to national and regional governing bodies to determine the scale of the project, based on how much are they willing to invest. The ongoing discussions include considerations of resources such as Nij Begun, a governmental initiative for economic recovery after phasing out the gas extraction in the area.

Infographic by Daniela Ivančáková
Infographic by Daniela Ivančáková

Possibilities and challenges

In terms of advantages, Sabatelli believes the AI factory would be a ‘big boost’ for RUG, but he also points out the amount of maintenance necessary for running such a project, which might create new job opportunities for scientific fields.


On the other hand, he sees some potential risks. “There is always this fear that even when you try to catch up, you will never be fast enough. By the time this hardware is there, and we are ready to use it, there might be something else, elsewhere, which is more powerful,” he says.


Another challenge lies in sustainability. “It would be great to keep in mind that it doesn’t come at zero cost, especially for the environment,” he emphasizes, talking about RUG’s own research on unconventional computing techniques. “From what I know, we're not there yet, so that might be the challenge.”


What is next?

The finalized application for the EuroHPC program should be on the table in May this year. The funding possibilities as well as the involvement of specific businesses and organizations are yet to be revealed.  



留言


bottom of page