The announcement of Stargate UK, a partnership between OpenAI, NVIDIA, and Nscale to deploy 31,000 GPUs within the UK, positions the United Kingdom as a leader in the debate over
It’s an inevitable comparison: how does the British plan compare to these European efforts?
Stargate UK: sovereignty with global partners
- Projected scale: starting with 8,000 GPUs in 2026, with ambitions to reach 31,000 GPUs in later phases.
- Location: multiple sites, including Cobalt Park in the AI Growth Zone in northeastern England.
- Tech partners: NVIDIA (with advanced GPUs, Grace Blackwell series with British-designed Arm architecture) and Nscale (local cloud operator).
- Primary use cases: critical public services, research, defense, finance, and regulated sectors.
- Social component: deployment of the OpenAI Academy to train 7.5 million workers in AI by 2030.
The British model relies on a transatlantic partnership with tech giants, but maintains national control over the physical infrastructure and access to models for sensitive uses.
GAIA-X: Europe’s vision for sovereign data
GAIA-X was launched in 2020 as a Franco-German project to create a European cloud data ecosystem based on transparency standards and digital sovereignty. Despite governance challenges and criticism over slow results, it remains the EU’s most ambitious framework.
- Goal: ensure that critical European data is processed on infrastructures compliant with the EU legal and value framework.
- Participation: over 300 entities from sectors such as energy, mobility, health, and public administration.
- Status: progressive deployment of vertical projects (GAIA-X Health, GAIA-X Mobility, etc.) and establishment of data federations.
Unlike Stargate UK, GAIA-X does not aim to deploy massive GPUs for AI training but instead seeks to ensure European data independence from non-EU cloud platforms.
France and Germany: data centers and supercomputing for AI
Both France and Germany have accelerated their sovereign AI agenda with direct investments in supercomputing infrastructure:
- France:
- Launch of Jean Zay and new HPC clusters dedicated to AI.
- Projects for green AI that reuse data center heat (such as the Data4 campus in Marcoussis).
- A national AI plan with €40 billion by 2030, including support for semiconductor startups and generative AI.
- Germany:
- JUPITER supercomputer, one of Europe’s first exascale systems, located at Forschungszentrum Jülich.
- Initiatives for climate-neutral data centers supporting European foundational model training.
- Debate over strict regulation (AI Act), aiming to balance innovation and rights protection.
United Kingdom vs. Europe: similarities and differences
| Aspect | Stargate UK (UK) | GAIA-X / EU (France, Germany, etc.) |
|---|---|---|
| GPU scale | Up to 31,000 GPUs (NVIDIA Blackwell) | Exascale supercomputing, not solely focused on GPUs |
| Partners | OpenAI + NVIDIA + Nscale (private/public) | Multi-company and multinational consortium |
| Focus | Sovereign computing for generative AI | Data sovereignty and cloud federation |
| Applications | Public services, defense, finance, research | Health, mobility, energy, public administration |
| Education | OpenAI Academy: 7.5 million workers trained | EU-funded digitization and digital skills programs | Political model | Bilateral UK–U.S. agreement | Community approach (27 countries, shared governance) |
Strategic reading
- Stargate UK emphasizes immediate computing power, partnering with US tech firms and aiming to attract critical applications.
- GAIA-X and the EU focus on data sovereignty and creating a regulated ecosystem where European services can flourish.
- France and Germany lead in supercomputing and regulated AI, with billion-euro investments aligned with the AI Act.
From a geopolitical perspective, the UK moves more quickly and pragmatically—relying on giants like NVIDIA and OpenAI—while the EU prefers slower processes but with a regulatory framework of its own that aims to protect its digital sovereignty in the long term.
Conclusion
The emergence of Stargate UK confirms that AI sovereignty is no longer an abstract concept but a strategic national objective. The UK opts for a flexible alliance with global leaders to secure computational muscle, while the EU concentrates on data control and regulatory standards.
Both paths reflect the same challenge: how to balance innovation, sovereignty, and competitiveness in the era of generative AI. The question isn’t whether it will be achieved but rather which model will be more effective in creating a self-sufficient ecosystem, reducing reliance on US and Chinese technology.
Frequently Asked Questions
What is the main difference between Stargate UK and GAIA-X?
Stargate UK focuses on sovereign computing with GPUs for generative AI, while GAIA-X aims for data sovereignty and cloud service federation in Europe.
How many GPUs does Stargate UK plan to deploy?
OpenAI and Nscale anticipate an initial capacity of 8,000 GPUs in 2026, expandable to 31,000 GPUs in later phases.
How will OpenAI support the UK workforce?
Through the OpenAI Academy, which will train 7.5 million workers in AI by 2030, in collaboration with local government and businesses.
Which European countries lead in sovereign AI?
France and Germany stand out with projects like Jean Zay, JUPITER, and their investment agenda, along with key participation in GAIA-X.