Nokia, four partners form Nordic-Baltic AI gigafactory consortium
Nokia, Nscale, Skeleton, Telia and Verda formed AIGF Northern to bid for a four-country Nordic-Baltic AI network, with submissions due in December 2026 and February 2027.
By Nathan Brooks
4 min read

What's News
- Five companies formed AIGF Northern on November 16, 2026.
- The planned network includes Oulu, Iru and Valmiera, with a Swedish site still unselected.
- Each EuroHPC facility must integrate at least three to four times the processor count in Europe’s most powerful AI factories.
- Lot 2 closes December 3, 2026; Lot 1 closes February 4, 2027.
- InvestAI targets €30 billion, including €20 billion in private funding and €10 billion in public sources.
By Nathan Brooks
Nokia, Nscale, Skeleton, Telia and Verda formed AI Gigafactory Northern (AIGF Northern) on November 16, 2026. Nokia said the five companies plan to bid to build and operate a federated, multi-site AI infrastructure network across the Nordic and Baltic region.
The consortium plans to enter the open consortium tender run by the European High Performance Computing Joint Undertaking (EuroHPC JU). The announcement is a proposed bid, not a tender award.
Nokia places the proposal within the EU-wide InvestAI initiative. The company says InvestAI aims to mobilize €30 billion for AI investment, comprising €20 billion in private-sector funding and €10 billion from EU and national public sources. That represents a two-to-one private-to-public funding split.
Nokia said the proposed network would connect facilities in Oulu, Finland; Iru, Estonia; and Valmiera, Latvia. The consortium has not selected the Swedish site. The three named locations plus that future Swedish site would spread capacity across four countries.
The consortium says distribution would let AIGF Northern draw on multiple sources of decarbonized energy, expand computing where capacity is available and reduce reliance on any single site or supplier. It also aims to avoid single points of failure and maintain continuity during disruptions.
Mikko Hautala, Nokia’s chief geopolitical and government relations officer, said the consortium unites connectivity, computing, cloud and energy capabilities needed for trusted, scalable infrastructure. “By linking strengths across borders, we can build a more resilient foundation for European innovation, competitiveness and technological sovereignty,” he said.
Nokia’s announcement sets out the division of work. Nokia would provide secure network infrastructure to link sites, move data reliably and support resilient operations. Nscale would bring cloud-platform and AI-infrastructure expertise for training and inference, while Verda would contribute AI-native infrastructure and cloud expertise for those workloads.
Nokia also said Telia and its subsidiary Telia Cygate would provide connectivity, data transmission, infrastructure operations and cloud capabilities. Skeleton would supply high-power energy-storage systems designed to improve data-center efficiency and stability while reducing strain on electricity grids.
VTT Technical Research Centre of Finland would support the project with quantum computing, semiconductor and photonics capabilities, Nokia said. KTH Royal Institute of Technology would contribute AI research, education and skills development. The consortium says the research partners would connect the infrastructure with Europe’s research and talent networks.
Nokia said the network would operate under European governance and combine large-scale AI computing with sovereign cloud and AI platform services. It would add high-capacity connectivity and decarbonized energy while serving businesses, researchers and public-sector organizations across Europe.
On its project site, AIGF Northern describes a technology-agnostic design for startups, researchers and next-generation companies. The model would coordinate AI delivery across Europe without dependence on a single provider.
The consortium says its regional case rests on reliable decarbonized energy, advanced digital networks, a cool climate that reduces cooling demand, heat-recovery potential and competitive energy costs. It says these conditions could deliver more computing capacity from existing grid connections, support lower-carbon operations and translate energy efficiency into more cost-competitive AI services.
Can the distributed model clear EuroHPC JU’s test?
EuroHPC JU requires qualifying gigafactories to integrate massive AI computing power, energy-efficient data centers and AI-driven automation. Those systems must optimize model training, inference and service deployment.
The scale test applies at each site. Each facility must integrate at least three to four times the number of the most advanced AI processors currently available in Europe’s most powerful AI factories. For AIGF Northern, that makes the threshold a site-by-site test rather than a combined network score.
Applicants also need adequate power-capacity infrastructure and best practices in energy efficiency, water efficiency and circularity. The call aims to provide sovereign AI computing for European researchers, public authorities, entrepreneurs, industries and small and medium-sized enterprises.
For AIGF Northern, the immediate task is to show that a distributed, cross-border design can satisfy requirements written for very large, energy-efficient facilities. EuroHPC JU will test whether that model can combine resilience and multiple energy supplies with the processor-scale and efficiency requirements.
The tender has separate lot deadlines. Lot 2 submissions are due December 3, 2026, while the Lot 1 deadline is February 4, 2027.
The extension for Lot 1 was intended to give member states time to complete national budget procedures and submit financial commitments. Those dates apply to the tender’s lots, not to a consortium-specific award schedule.
What changes for European AI users?
If AIGF Northern wins and meets the requirements, European users would gain access to a proposed sovereign computing and cloud network spanning a planned four-country Nordic-Baltic footprint. The consortium says the distributed model could reduce single-provider dependence and improve service continuity. It could also give startups, researchers and companies more locations in which to run training and inference workloads.
Yet AIGF Northern remains a proposed bid, not an awarded project. The consortium must show that its federated design can meet the processor, energy and water standards without losing the resilience advantage.
The next milestones are the two tender deadlines. EuroHPC JU’s processor and efficiency standards will decide whether AIGF Northern’s planned four-country model advances beyond proposal.
Source: unite.ai
More from Nathan Brooks
Show full bio
News editor covering regulation, policy and their impact on entrepreneurship, startups and venture funding at Business Bearings.
2,232 articles