Spain Pitches AI Chip Hub

Europe is good at writing AI rules. The harder question is whether it can run the machines on which AI depends. If the computing power sits in the United States or Asia, European regulation still leaves a practical weakness: the switch, the chips and the cloud contract are somewhere else.

Spain is now making one of the more concrete European attempts to close that gap. The bid is not another policy paper from Brussels. It is a proposal built around land, electricity, cooling systems and a small town in Tarragona province.

A factory for calculations

A normal data centre is built to run many different jobs at once: company databases, streaming services, websites, payment systems. An AI gigafactory is different. It links tens of thousands of specialised chips so they can work together on one huge calculation, such as training a language model or a medical imaging system.

That only works if the whole site is designed around the chips. Storage, cooling and networking have to move quickly enough so the processors are not left waiting for data (Hammerspace, Schneider Electric). In Maltese terms, this is not the digital equivalent of renting a few more offices in Mriehel. It is closer to building industrial infrastructure for a new kind of production.

The EU wants to fund up to five such gigafactories across Europe, each with around 100,000 advanced AI chips (Data Center Dynamics). The call for bids opens this month, with a decision expected by late 2026 (elDiario.es).

Spain's proposal is centred on Móra la Nova, a small town in Tarragona province, with a secondary site near Madrid (Europa Press).

Why the Spanish state wants to co-own it

The ownership plan shows how Madrid sees the project. A state investment vehicle would hold roughly 48%. Telefónica, construction group ACS and Santander would each take about 16%, while quantum-computing firm Multiverse Computing would hold 4% (La Vanguardia, 20minutos).

That is not the structure of an ordinary private cloud campus. It is the structure of strategic infrastructure, with the state close to control and the private sector supplying telecoms capacity, construction expertise and finance. For Malta, which has built whole economic sectors around regulated digital services, from iGaming to fintech, the lesson is familiar: the regulatory licence matters, but the pipes and capital behind it matter just as much.

The money also needs some care. Public commitments that can be identified so far amount to about €1 billion, combining Spanish state funds and a contribution to the EU's joint supercomputing programme (esmartcity). The €5 billion figure widely reported is the hoped-for total investment if the bid succeeds, not money already committed (elDiario.es).

The real bottleneck is electricity

An AI gigafactory is an industrial plant. It needs land, power and water before it needs slogans about sovereignty. At 100,000 chips, the electricity demand alone reaches the level of a small city.

Móra la Nova's local substation currently delivers 50MW. The plan is to expand that towards 125MW by 2030 (20minutos).

Ireland is the warning case. Data centres became such a large part of national electricity demand that regulators now test each new grid connection against local capacity and stress on the wider system (CRU). Spain will need transmission upgrades, stronger substation capacity and agreements to reduce or shift demand when the grid is under pressure. Those arrangements have to be ready before the chips arrive.

Who actually gets to use it

The strongest argument for doing this comes from Portugal's parallel bid. The promise there is that gigafactory capacity would be available to small businesses, startups, researchers and public services (Government of Portugal).

That is where the policy becomes practical. A hospital fine-tuning a medical AI model, or a university training a language system, could use European computing power without sending sensitive patient or research data through a non-European cloud provider. For a small member state like Malta, this matters because access is often the difference between participating in a European technology shift and merely buying services from whoever already dominates the market.

France shows both the potential and the limit. It attracted €93 billion in foreign AI-related investment pledges at its Choose France summit, helped by the availability of nuclear-powered electricity (Le Monde). Mistral AI's new data centre near Paris runs on 13,800 Nvidia chips, all American-designed (Le Nouvel Economiste).

Europe can decide where AI systems run, who can access them and what rules govern the data. Independence from US-designed chips is still a distant prospect (EU AI Policy Monitor).

Spain has not won the bid. If it is selected, operations could begin around late 2028 (La Vanguardia). Much of Europe's AI funding still exists as programmes and financing vehicles, not as machines that researchers can book time on (Thorsten Meyer AI).

The real test is whether Móra la Nova's substation, grid connection and cooling systems can be built before the AI race moves on. AI sovereignty will be decided as much by concrete and electricity as by regulation.