Americans’ fervor for Europe’s abrupt announcement of an AI Gigafactory plan may just be a question of sexy marketing, but when European politicians dared to grip the Tesla steering wheel on the future, it perked up American ears.
The European Commission has been inundated with companies keen to build AI gigafactories across Europe.
Proposals for new AI gigafactories had poured into Brussels, with 76 expressions of interest filed, for 60 sites in 16 countries in the EU.
This strong interest has certainly emboldened officials, who see it as endorsement of their approach to try to carve Europe’s place as a heavyweight in the global AI race. But these aren’t symbolic gestures — ideas have been floated by major industry players.
Big European data center operators, telecom giants, energy providers and global tech companies have all said they want a piece of Europe’s AI revival.
Commission officials still refuse to reveal the exact companies involved – citing commercial confidentiality – but those in the industry are already speculating over what alliances may soon shape up among Europe’s top tech players.
Between them, these companies plan to buy at least three million GPUs, or enough computing power to train the most advanced AI models that could compete with or even surpass what’s available today.
AI Gigafactories: It’s Not Just About The Size Of The Computer
So what, exactly, is an AI gigafactory? It’s the digital analog to a huge manufacturing plant. But unlike creating physical goods, these factories will be churning out the next generation of AIs.
Rather, we are talking about computer environments on a scale not seen before. The goal of these facilities is to provide Europe with the computing infrastructure it requires to develop AI capabilities that are not entirely dependent on American or Chinese technology.
Gigafactories are an extension of the EU’s current AI strategy. They will complement the groundwork being laid by Europe’s excellent EuroHPC supercomputing infrastructure, with one important difference — they are dedicated exclusively to AI, and they have a strong commercial basis.
You don’t need to be a follower of tech news to understand what’s at stake with the debate over online piracy. A race to build advanced AI infrastructure has intensified wildly over the past 18 months by the spending of billions of dollars by American tech giants on data centers and custom chips. And China has not slowed its aggressive push in the sector.
Europe has frequently been accused of moving too slowly or getting mired in bureaucracy. But this effort is evidence that Brussels is also now setting out to stake its own claim. What the Commission is trying to do is create a climate where European AI can flourish by coordinating investments between member states and public and private resources.
And it’s not just a question of computing power. It’s about creating a whole ecosystem where hardware, software, data and talent can be together. That is what makes the idea of AI gigafactories so potentially transformative.
The Energy Question
And then there’s the elephant in the room: energy. Training cutting-edge AI models takes gargantuan amounts of electricity, and the act of bolting down millions of power-hungry GPUs to Europe’s computing infrastructure has all but everyone fretting about sustainability.
It’s believed that many of the proposals include creative cooling solutions and so-called backsliding agreements with renewable energy providers. One of the consortium is understood to be looking at a site in northern Sweden which would be powered by 100 per cent hydropower, and reliant on the region’s naturally cool climate for cooling.
What’s Next?
The Commission will subsequently consult all the respondents to determine the next step of the project. There will be no formal call to establish these gigafactories before the end of 2025, while the process will be coordinated by the EuroHPC Joint Undertaking.
That timeline might feel frustratingly slow to some in the industry, but it is a testament to the complexity of orchestrating such ambitious projects across countries and regulatory systems.
What This Means for the Ordinary European
The effects may not seem apparent to the layperson right away. But successful, they could one day end up everywhere — from health care diagnostics and climate modeling, to customer service bots and next-generation gaming experiences.
The true issue is whether Europe can act quickly enough. In the international AI race, finishing second or third might not just be a matter of pride — it could also relegate you to forever importing foreign AI technology rather than developing home-grown capabilities.
