The European Union has announced a $30 billion initiative to construct a network of high-capacity AI data centers, aiming to strengthen its position in the global artificial intelligence market.
The EU has allocated €10 billion (approximately $11.8 billion) to establish 13 AI data centers, with an additional €20 billion earmarked as initial funding for a network of gigawatt-class AI facilities. This ambitious project represents one of the world’s largest publicly funded AI initiatives and marks a significant effort by the bloc to close the technological gap with the United States and China. The project has already generated substantial interest, attracting 76 expressions of interest from 16 member states covering 60 potential locations across Europe.
Implementation is proceeding rapidly, with the first AI factory expected to launch in the coming weeks and a large-scale project in Munich planned for early September. Each gigawatt data center is projected to cost between €3 to €5 billion and will deliver computational power far exceeding current AI facilities. According to estimates by UBS, these facilities could potentially support over 100,000 advanced AI GPUs per site. For comparison, xAI’s Colossus super cluster consumes about 150 MW of power when equipped with 100,000 H100 GPUs, suggesting that a gigawatt facility might host significantly more processing units, potentially up to 300,000 Blackwell Ultra processors.
Henna Virkkunen, European Commission executive vice president for technology policy, highlighted Europe’s strong talent base, noting that the continent reportedly has 30% more AI researchers per capita than the United States. However, limited access to computational resources has hindered European AI development, with these massive data centers intended to address this critical bottleneck and stimulate the EU’s AI sector.
Despite the substantial investment and enthusiasm, significant challenges remain. Bertin Martens of Bruegel pointed out uncertainties regarding the total public sector investment commitment and the technical specifications of the planned data centers. While the EU has access to advanced AI accelerators through a trade agreement with the U.S., Martens emphasized that acquiring hardware is merely the first step in a complex process.
Perhaps the most pressing concern is power supply. Each gigawatt-class data center will require enormous electrical capacity, but developing such generation infrastructure typically takes years. Martens cautioned that Europe’s current grid infrastructure may not be prepared to support these high loads, necessitating massive upgrades that could extend the project timeline significantly. Additionally, ensuring the facilities operate profitably will require developing sustainable business models and securing interest from private companies.
