Companies

OpenAI Announces Stargate UK With NVIDIA and Nscale

OpenAI, NVIDIA and Nscale will build sovereign UK compute capacity — up to 8,000 GPUs by Q1 2026, scaling to 31,000 — plus an AI skilling push targeting 7.5 million workers.

Introducing Stargate UK
Introducing Stargate UKAI-generated
By Sophie Lindqvist3 min read

Updated

Why it matters

  • OpenAI will explore offtake of up to 8,000 GPUs in Q1 2026, with potential to scale to 31,000 GPUs over time.
  • Stargate UK sites include Cobalt Park, part of the newly designated AI Growth Zone in the North East.
  • OpenAI is bringing its Academy to the UK to support the Government's goal of upskilling 7.5 million workers by 2030.

OpenAI will explore offtake of up to 8,000 GPUs in Q1 2026, with potential to scale to 31,000 GPUs over time, under a new AI infrastructure partnership announced today called Stargate UK.

The initiative, built with NVIDIA and UK-based compute provider Nscale, is designed to run OpenAI's models on local computing power inside the UK — aimed at specialist use cases where jurisdiction matters. OpenAI named critical public services, regulated industries like finance, research projects, and national security partnerships as target workloads.

The deal follows a memorandum of understanding OpenAI signed with the UK Government in July 2025 to explore the country's infrastructure priorities and accelerate AI adoption. Stargate UK is the latest rollout of OpenAI for Countries, the company's program for governments seeking sovereign AI capabilities, and OpenAI says it supports the UK's national AI Opportunities Action Plan.

The stakes are straightforward. Sovereign compute has become a strategic asset for governments that want the benefits of frontier AI models without sending sensitive workloads through foreign jurisdictions. Stargate UK positions Britain to keep regulated and security-related AI processing on domestic soil, while giving OpenAI a beachhead in one of its most important markets.

Nscale plans to significantly expand its planned UK capacity for the project. Stargate UK is expected to be based across multiple sites, including Cobalt Park, which will form part of the newly designated AI Growth Zone in the North East of England.

NVIDIA will supply its most advanced GPUs for the build-out. British chip designer Arm contributes semiconductor designs to NVIDIA's latest Grace Blackwell series of chips, a detail OpenAI highlighted as evidence of UK-US collaboration running through the hardware stack itself.

OpenAI CEO Sam Altman framed the partnership as building on Britain's existing strengths. "The UK has been a longstanding pioneer of AI, and is now home to world-class researchers, millions of ChatGPT users, and a government that quickly recognized the potential of this technology," Altman said. "Stargate UK builds on this foundation to help accelerate scientific breakthroughs, improve productivity, and drive economic growth. This partnership reflects our shared vision that with the right infrastructure in place, AI can expand opportunity for people and businesses across the UK."

Hardware is only half of the announcement. OpenAI also said it will bring the OpenAI Academy — its AI literacy hub covering AI fundamentals, available tools, and best practices — to the UK. The program backs the Government's stated ambition to upskill 7.5 million workers by 2030. OpenAI said it will partner with local business leaders to design flexible, customizable learning experiences, with the stated aim of preventing regional gaps in AI adoption, and will share more details with the UK Government in due course.

The UK matters commercially to OpenAI beyond infrastructure. The company says the UK ranks among its top five markets globally for paid subscribers and API developers. Its customer base there spans NatWest, Virgin Atlantic, London-based video AI unicorn Synthesia, and Oxford University, alongside everyday users, developers, startups, and institutions.

The announcement also deepens the US-UK technology partnership at a moment when national compute capacity has become a proxy for economic competitiveness. With the first GPU offtake targeted for Q1 2026, the test now shifts to execution: whether Nscale and its partners can deliver capacity on schedule, and whether UK enterprises and public bodies move fast enough to use it.

Source: OpenAI News

Share this article:

More from Sophie Lindqvist

Sophie Lindqvist

Show full bio

Staff writer covering marketplaces and e-commerce at AI In Context.

115 articles

Related articles

  1. OpenAI Launches 'OpenAI for Countries' to Export Stargate-Style AI Infrastructure
  2. OpenAI Announces Stargate Norway, a 100,000-GPU Data Center Near Narvik
  3. OpenAI Launches 'OpenAI for India' With Tata Data Center Deal
  4. AWS and OpenAI Sign $38 Billion Multi-Year Compute Deal
  5. OpenAI Launches Australia Program With Sovereign AI Campus

« Previous articleNext article »