OpenAI has announced its plans to open a data centre in Narvik, Norway, as part of Stargate Norway, which will see the facility deliver 100,000 NVIDIA GPUs by the end of 2026.
This will mark its first data centre in Europe and belongs to its OpenAI for Countries programme, which seeks to work collaboratively with companies in building their AI infrastructure.
Several companies have existing strategies, some of which date back into 2017 in what perhaps could be regarded as a prescient outlook on how AI would evolve: ChatGPT, OpenAI’s tool, was launched in 2022. Among them was the UAE, which published an AI strategy in 2017 which outlined its mission of becoming a world leader in AI by 2031, and created a federal council for AI and blockchain technology which was headed by Omar Sultan Al Olama, the world’s first minister for AI.
The UK government unveiled its AI Opportunities Action Plan in January 2025 and in more recent news, signed a Memorandum of Understanding (MoU) with OpenAI to accelerate adoption, drive economic growth, and improve public services with the use of AI. Elsewhere, the European Commission announced its InvestAI initiative which oversaw the launch of a €200 billion investment package, with €20 billion set aside for building four AI gigafactories.
And Egypt’s National AI Strategy plans to raise the contribution of the ICT sector to GDP to 7.7%. France’s France 2030 programme’s AI strategy commits €1.5 billion to build sovereign compute infrastructure, which is an important point: many of these countries are looking to establish AI sovereignty and not rely on regions overseas.
However, a fundamental pillar of sovereign AI is infrastructure. Companies need the proper AI infrastructure, in the form of data centres with advanced computing capabilities, in order to compete on the world stage. This is where OpenAI hopes to help.
Other pillars that are important to sovereign AI are having a skilled workforce in place, investing in research and development, and creating an environment conducive for AI-driven businesses and applications. It also depends on having a clearly defined regulatory and ethical framework, and international cooperation
Investments in AI infrastructure
The Stargate project was initially launched as an investment in the US’ AI infrastructure and announced in January 2025. In the announcement, OpenAI positioned the infrastructure as the “backbone” of future economic growth and national development.
The aim of the project, the company noted, was to spread democratic AI, which allows people to choose how they work with and direct AI independently of government control, and fostering a free market.
The partners on Stargate Norway are Nscale, an AI infrastructure provider, and Aker, a Norwegian engineering firm. The project will be a joint venture between both businesses who will provide access to Norway’s AI ecosystem in a bid to ensure that startups and scientific researchers will benefit from the additional compute.
OpenAI currently relies on NVIDIA’s chips and although it confirmed it was exploring Google’s Tensor Processing Units (TPUs) to see if they can sufficiently meet demand for AI computing, it said it had no current plans to deploy them at scale.
NVIDIA’s GPUs are fairly dominant in the market. Although figures vary – IoT Analytics, for instance, reported that NVIDIA commanded 92% of the global data centre GPU market in 2024 while TechInsights reported in 2023 that NVIDIA held 63% share – average estimates place its market share at around 80%.
NVIDIA has real reason to invest in the data centre business. That portion of the business made up almost 90% of its revenue in Q4 2024. There are other reasons why NVIDIA enjoys a dominant market position: one being that its GPU-powered solutions are widely available through major Cloud service providers, including AWS, Google Cloud, and AI companies like OpenAI, but also Anthropic, and Meta.
In the announcement of Stargate Norway, OpenAI called it “one of the most ambitious AI infrastructure investments” in Europe, although it is not alone in making these investments. Amazon Web Services (AWS) is planning $100 billion in capital expenditure during 2025, predominantly for hyperscale data expansion and Trainium2 AI chip development, while Microsoft committed $80 billion in 2025 with more than half allocated to AI-optimised data centres in the US.
It’s clear that countries that have recognised leading in AI development and adoption will position them competitively and attract business investment, and companies like OpenAI that provide the ‘picks’ for this modern-day gold rush, will arguably benefit the most.
