Polarise is preparing to enter the European Union’s tender for a new generation of AI gigafactories, positioning the company inside a much larger debate over who controls the continent’s compute infrastructure. The European AI infrastructure programme could support up to seven gigafactories, backed by as much as €10 billion in EU and national funding. The initiative aims to mobilise at least €20 billion in private investment, turning public capital into a catalyst for large-scale sovereign AI capacity. For Polarise, the opportunity is less about securing a single large facility and more about building a connected infrastructure model that can operate across European locations.

The company intends to participate in the tender through a consortium, bringing its existing infrastructure footprint and sovereign AI cloud capabilities into a broader European bid. That strategy reflects Polarise’s view that future AI infrastructure does not necessarily need to concentrate compute inside one enormous campus. Instead, geographically distributed facilities could function as a coordinated layer of European compute, giving operators more flexibility around capacity, location and infrastructure deployment. This approach sits at the centre of CEO Michel Boutouil’s argument for what Polarise calls a connected “Gigagrid” architecture.

Michel Boutouil, CEO of Polarise, said, This tender provides exactly the impetus we have long been calling for: a connected ‘Gigagrid’ architecture instead of individual large-scale sites. Looking ahead, this means that a single large site will no longer necessarily be required – connected sites with a shared compute-equivalent capacity can also contribute to a powerful European AI infrastructure. Europe now has a narrow window of opportunity to build sovereign AI capacity. That is why we are contributing our experience and sites to a corresponding consortium and shaping European sovereignty together with strong partners.

Existing Sites Give Polarise A Head Start

Polarise enters the tender discussion with infrastructure that already exists rather than a plan built entirely around future development. The company points to its operational AI Factory at Munich’s Tucherpark as evidence that it can develop and run modern AI infrastructure at commercial scale. It has a further site in Oslo, expanding the geographic footprint available to a potential consortium. That operational experience gives Polarise a more tangible starting position as Europe evaluates proposals for its next wave of AI capacity.

The company says it has additional prepared locations that could support the infrastructure strategy, including two sites in Bavaria and a potential site in North Rhine-Westphalia. Those locations matter because the Gigagrid concept depends on more than raw accelerator capacity. A distributed architecture needs sites that can be developed, connected and operated as part of a coordinated compute environment rather than treated as isolated data-centre projects. Polarise’s existing pipeline therefore gives its consortium ambitions a physical dimension that extends beyond its current operational footprint.

The Munich and Oslo projects further establish the company’s experience across different infrastructure environments. That matters in a European market where AI compute demand is colliding with questions around energy availability, network connectivity, land, regulation and national strategic interests. A multi-site architecture can potentially distribute those constraints instead of forcing every requirement into one location. As a result, Polarise is presenting its existing sites as strategic infrastructure assets rather than simply real-estate options for a future tender.

Polarise Brings AI Software Into The Infrastructure Bid

Polarise’s proposition extends beyond physical compute capacity, with the company already developing software and services intended to sit above the infrastructure layer. Its Polarise Core platform provides managed ML-Ops and development capabilities through a platform-as-a-service model. The company says these capabilities already cover functions that the EU tender framework anticipates in later stages of development. That could allow Polarise to approach the programme with an integrated infrastructure-and-platform proposition rather than relying solely on hardware deployment.

The company’s Polarise Drive platform adds another layer to that stack through integrated AI services. Those services include Model-as-a-Service, inferencing and APIs, with a particular focus on open-source models. The combination gives Polarise a route from underlying compute to usable AI services for developers and organisations that need access to models without building the entire operational stack themselves. For an EU programme centred on sovereign AI capacity, that distinction could become important because infrastructure alone does not automatically translate into accessible European AI capability.

Polarise’s software strategy reflects a broader shift in how AI infrastructure providers compete. The market is moving beyond the question of who can install the largest number of accelerators and toward who can make those resources usable, manageable and commercially accessible. Polarise is building around that operational layer with Core and Drive, creating a proposition that links infrastructure with model development and inference services. That gives the company a potentially differentiated role inside a consortium where physical sites, compute resources and software services need to function as one system.

Polarise Positions The Tender As An Accelerator

Polarise’s participation would give the EU tender a candidate with operational infrastructure, prepared sites and an existing AI cloud platform already moving beyond the early stages of deployment. Its strategy combines physical capacity with managed AI infrastructure and services, while its Gigagrid proposal challenges the assumption that sovereign compute must reside within a small number of massive facilities. The company will now need to convert that proposition into a consortium structure capable of meeting the technical and strategic demands of a programme worth billions of euros. The outcome could help define whether Europe’s next AI infrastructure cycle favours centralised gigafactories or a more connected network of sovereign compute sites.

For Polarise, the underlying strategy remains intact regardless of the tender outcome. The company is continuing its site and platform development while using the EU initiative as an opportunity to accelerate the scale of those efforts. Its existing Munich and Oslo infrastructure, planned sites across Germany and its Core and Drive platforms provide the foundation for that expansion. The next phase will depend on the partners Polarise brings together and how convincingly the company can turn its Gigagrid vision into a deployable European AI infrastructure model.