Nscale Rises Rapidly to Prominence in AI Infrastructure, Drawing Major Backing from Nvidia and Partners

Two years ago Nscale was an obscure London startup with no external funding and no public presence. After emerging from stealth last year, the company disclosed in December that it had closed a $155 million Series A round. In a matter of months Nscale has surged to the forefront of the global artificial intelligence infrastructure market and now counts nearly $700 million in new capital from Nvidia among its backers.

This week the company announced a series of high-profile partnerships and commercial commitments. Nscale was named as an AI infrastructure partner for Nvidia, Microsoft and OpenAI as each expands its U.K. footprint. The firm also revealed a five-year, $6.2 billion agreement with Microsoft and Aker to build hyperscale AI infrastructure in Norway. OpenAI previously disclosed plans for a Stargate-branded AI data centre in Norway, for which Nscale has pledged $1 billion and set an objective of installing 100,000 Nvidia GPUs at the site by 2027.

Nscale’s rapid ascent traces back to its origins inside Arkon Energy, which was founded to supply infrastructure for cryptocurrency mining. The business was spun out to service the soaring demand for data centres optimised for AI workloads. Like CoreWeave — itself a former crypto-focused operator that this year listed publicly and now has a large market valuation — Nscale combines power, rack space and large GPU inventories with software to offer integrated AI infrastructure services.

Founder and CEO Josh Payne told media in late July that Europe faces two material challenges: insufficient compute capacity and a fragmented supplier market. He argued that large-scale infrastructure projects are required to meet enterprise demand and to enable productivity and economic benefit across the region. Payne described the Microsoft–Aker agreement as a major win for Europe’s locally owned AI infrastructure ambitions.

Microsoft led the U.K. announcements with a pledge of $15.5 billion in new investment in computing equipment and disclosed plans to work with Nscale to build what will become the U.K.’s largest supercomputer in Loughton, Essex. The site is slated initially to host 23,040 Nvidia Blackwell GPUs expected to arrive in the first quarter of 2027 and to provide 50 megawatts of AI capacity, expandable to 90 megawatts. Microsoft President Brad Smith said the scale of the capital deployment reflects strong commercial commitment to the project.

OpenAI will deploy 8,000 GPUs in the first phase of its U.K. Stargate roll-out and reserves the option to expand capacity to roughly 31,000 GPUs. Stargate U.K. is planned to operate across multiple sites, including Cobalt Park in Newcastle. Nvidia also announced a broader investment package of up to £11 billion ($15 billion) with Nscale and CoreWeave to accelerate U.K. AI infrastructure. Separately, Nvidia CEO Jensen Huang disclosed a £500 million ($683 million) equity investment in Nscale.

Nscale’s commercial momentum accelerated after its public debut. Following its exit from stealth in May of the prior year, the company formed a partnership with Open Innovation AI to deploy 30,000 GPUs and acquired Kontena, a high-performance computing data-centre specialist founded in 2018. The firm later agreed with Singtel to offer GPU-as-a-Service for customers across Europe and Southeast Asia. While Nscale initially used GPUs from AMD, its current offerings prominently feature Nvidia hardware.

The $155 million Series A was led by Sandton Capital Partners with participation from Kestrel0x1, Blue Sky Capital Managers and Florence Capital. At the time of that raise Nscale reported expanding its AI data-centre pipeline to 1.3 gigawatts from 300 megawatts a year earlier and set a target of having 350,000 GPUs operating by the end of 2027. In related comments, Sandton co-founder Rael Nurick described Nscale’s vertically integrated model as building a hyperscale platform for AI at scale.

Industry observers note the capital intensity of large AI data centres. CoreWeave, for example, had amassed significant debt financing and sizable equity before its IPO, and recently completed large bond offerings to fund expansion. Nscale advanced efforts to secure private credit earlier this year, reportedly seeking $1.8 billion in a financing led by Goldman Sachs.

Payne has publicly framed the company’s strategy around increasing GPU density and scaling infrastructure rapidly. In a December video accompanying the equity raise he set ambitious near-term deployment goals, including 50,000 GPUs by the end of 2025 and 150,000 by the end of the following year, alongside incremental capacity and GPU additions in Norway and other markets.

Nscale’s trajectory reflects a broader industry shift: operators that combined early crypto-era GPU deployments with data-centre expertise have been able to pivot into the AI infrastructure market and secure strategic relationships with hyperscalers and major AI vendors. For Nscale, the challenge ahead will be executing an aggressive build-out at a time when such projects require substantial capital, logistical coordination and long-term commercial commitments from large customers.