A four-country AI network could give European channel partners another source of computing capacity. Whether they can build a business around it remains unclear.
Nokia, Telia, Nscale, Verda, and Skeleton announced AI Gigafactory Northern on Nov. 16. The consortium seeks to develop large-scale AI infrastructure under European governance, giving businesses and public institutions greater access to computing resources within the region.
The consortium intends to bid under the EU’s AI Gigafactories program. Its announcement does not set out commercial access terms for outside MSPs, resellers, or systems integrators.
Proposed AI network would span four countries
AI Gigafactory Northern (AIGF Northern) would link computing facilities across four countries. Proposed locations include Oulu in Finland, Iru in Estonia, and Valmiera in Latvia, with a Swedish site still to be determined.
Designed as a federated system instead of one large campus, the network would provide capacity for AI training and inference workloads.
According to Nokia, the founding companies would contribute the following technologies and expertise.
| Company | Planned contribution |
| Nokia | High-performance networking equipment to connect the facilities and handle data transmission between sites. |
| Telia and Telia Cygate | Telecommunications connectivity, infrastructure operations, cloud solutions, and systems integration expertise. |
| Nscale and Verda | AI computing capacity and cloud platforms to run customer workloads. |
| Skeleton Technologies | High-power systems intended to improve data center stability and efficiency while reducing strain on electricity grids. |
The consortium cites Northern Europe’s cooler climate and access to lower-carbon electricity as advantages for the proposed locations. It also says distributing capacity across countries would reduce reliance on a single site as AI data center development across Europe increases demand for electricity.
Research support will come from VTT Technical Research Centre of Finland and Sweden’s KTH Royal Institute of Technology. Their contributions include advanced computing research, AI expertise, and skills development to support the initiative.
Consortium plans bid under EU gigafactory program
Under the European Commission’s AI Gigafactories initiative, the EU aims to establish up to seven facilities across the continent. Launched in July, the program targets more than €30 billion in public and private investment.
Up to €10 billion could come from EU and national funding, with at least €20 billion expected from private investors. Those targets cover the wider program, not money already awarded to AIGF Northern.
Selected groups would establish and operate the gigafactories, with participating public authorities purchasing access to their computing capacity.
According to the EuroHPC Joint Undertaking tender, qualifying projects must have at least three to four times as many advanced AI processors as Europe’s most powerful existing AI factories. Bidders must also demonstrate adequate power infrastructure and meet requirements covering energy and water efficiency.
Large-scale bids close Dec. 3, 2026, and medium-scale submissions are due Feb. 4, 2027.
Four-country AI network creates a different challenge for partners
Connecting four facilities would not automatically turn them into one enormous computing cluster. The announcement does not explain whether individual AI training jobs could use processors spread across multiple countries.
For systems integrators advising customers on AI infrastructure deployments, the distinction affects which workloads they can recommend.
Training a large model requires processors to exchange information rapidly, often favoring computing capacity concentrated at one site. Running AI assistants for customers across several European markets is a different case, where regional inference services may be more useful.
Cloud-focused MSPs could build services around regional deployment, but only if the operator allows them to select locations and move workloads when demand or available capacity changes.
Winning the EU bid would not automatically establish a resale market for the gigafactory’s computing resources. Outside providers would need commercial access that allows them to reserve capacity and offer managed AI services to their customers.
For partners evaluating a potential service offering, the next step is to establish what they could sell, control, and support. Before recommending the proposed network, they should seek details on:
- Commercial access: Whether outside providers could reserve and resell capacity, and under what pricing and commitment terms.
- Workload placement: Whether customers could select processing locations and control transfers between sites.
- Service responsibility: Which organization would handle outages, customer support, and service-level commitments.
- Deployment requirements: Which workloads could run at individual sites, what migration would involve, and how performance would be measured.
Those answers would help partners assess margins, support costs, and suitability for customer workloads. The proposed capacity creates a reason to investigate; a viable managed service would still depend on the operating model and commercial terms.
More on AI: Storage modernization could help businesses reduce power consumption while making room for more AI workloads. Explore the opportunities for systems integrators.


