NSCALE, an AI infrastructure provider at its UK headquarters.
Courtesy: nscale
Two years ago, Nscale was a new startup in the UK, and had not officially announced its external funding or its existence.
Last year, the London-based company came out of stealth, but raised Series A funding in December, revealing a total of $155 million.
Currently, NSCALE is discovering it is at the heart of the actions of artificial intelligence, the hottest market on the planet. And it has nearly $700 million fresh capital from nvidiathe most valuable company in the world.
In a press release on Tuesday, Nscale was designated as an AI infrastructure partner nvidia, Microsoft Openai said it has signed a five-year $6.2 billion agreement with Microsoft and Aker to expand its buildout at NSCALE in the UK, and has developed a “hyperscale AI infrastructure” in Europe, particularly in Norway, where Aker is headquartered.
Openai created a previous headline on NSCALE and announced plans for the Norwegian data centre for the Stargate brand AI data centre in July. NSCALE agreed to commit $1 billion to the project prior to 2027 with the goal of acquiring 100,000 NVIDIA Graphic Processing Units (GPUs) on its site.
It’s a very quick rise for a company that wasn’t even able to even begin a generation AI boom with the launch of ChatGPT in late 2022. At the time, NSCALE was part of Arkon Energy, which was founded a year ago to provide infrastructure for cryptocurrency mining. NSCALE was spun to deal with the rising demand for data centers that can handle AI workloads.
Like coreweaveNSCALE, which was released this year and now boasts a market capitalization of $58 billion, combines data center space, power and many GPUs with its own software to provide end-to-end services to its AI infrastructure.
CoreWeave, which provides infrastructure to Microsoft, GoogleNvidia and Openai also have roots in cryptography. Founded in 2017, the company built the first fleet of Nvidia GPUs for Ethereum mining before pivoting into AI.
Nscale did not respond to requests for comment following this week’s announcement, but previously founded Arkon CEO Josh Payne told CNBC in late July that the company was targeting two major issues in Europe. One lacks sufficient computing power, and the other is a “very fragmented market.”
“What the continent needs is for a large AI infrastructure project to deploy Compute (Power),” Payne said after making an announcement at Norway Buildout’s Openai. “The ecosystem can consume from projects to build AI products, creating productivity growth and economic benefits.”
Payne wrote in a LinkedIn post Wednesday that the agreement with Microsoft and Aker is “a big victory for European-owned AI infrastructure.”
Europe is pushing for the concept of “sovereign AI” and requires data centers and AI workloads to be placed and processed in European soils. NSCALE quickly emerged as a key player in the UK’s bid to evolve into a global leader in AI. In January, the UK submitted an AI “plan of action” to promise to reduce bureaucracy to help the domestic AI sector thrive.

Although NSCALE is committed to the European market, many of its early partners are large American AI vendors. They timing the announcement Tuesday for President Donald Trump to visit the UK
On Wednesday, Trump visited Windsor Castle to meet King Charles, Queen Camilla and other royal members. His trip comes at a controversial moment for British Prime Minister Kiel Starmer. Kiel Starmer has been under pressure to bring stability to the country after Deputy Prime Minister Angela Rayner’s departure over the House tax scandal and major cabinet reshuffle.
Microsoft has spotted the UK’s announcement and has made a new $15.5 billion investment in computing equipment. The software giant said it plans to work with Nscale to build what will become the UK’s biggest supercomputer in Loughton, a suburban town in Essex’s UK county.
According to a statement from NSCALE, the site will initially house 23,040 NVIDIA BLACKWELL GPUs in the first quarter of 2027.
“No one can make such capital investments unless someone has already committed to spending money once the job is done, and that’s the role we play,” Microsoft President Brad Smith said Tuesday, saying the deal represents a major vote of trust in NSCALE.
Openai said it will launch the UK version of Stargate through its partnership with Nscale and Nvidia. Openai will deploy 8,000 GPUs in the first phase of the project early next year, and will roll out the option to expand its capacity to approximately 31,000 GPUs over time.
Stargate UK operates on many sites across the country. One of the early sites was Cobalt Park, an industrial nation in northern Newcastle. Stargate was originally generated in the US in January as part of President Trump’s efforts to drive investment in AI infrastructure.
Nvidia CEO Jensen Huang will be attending the “AI Race Winning” Summit held in Washington, DC, USA on July 23, 2025.
Kent Nishimura | Reuters
Nvidia’s announcement on Tuesday included investments of up to £11 billion ($15 billion) at NSCALE. coreweave Enhance the UK’s AI infrastructure.
Nvidia CEO Jensen Huang separately revealed on Wednesday that Chipmaker had made a £500 million ($683 million) equity investment in Nscale.
“We are confident that Nscale could become the national champion of UK AI infrastructure,” Huang told journalists at a press conference in London.
Nick Patience, the AI practice lead at Futurum Group, told CNBC that Nscale was “a long slog that we have approved that something has to be done to build a government AI infrastructure here,” and that Nscale has been a long slog.
Rapid growth
After closing Stealth in May last year, NSCALE’s first public announcement came two months later, with the company partnering with UAE’s Open Innovation AI to deploy 30,000 GPUs. Around the same time, Nscale said it was acquiring Kontena, which was founded in 2018 and specializes in high-performance computing data centers.
The following month, NSCALE announced an agreement with Asian telecommunications company Singtel to provide “GPU-As-A-Service (GPUAAS)” to serve customers in Europe and Southeast Asia. Initially, NSCALE’s infrastructure was GPU-dependent Advanced Micro Devices. Today, startups promote a variety of products from market leader Nvidia.
NSCALE’s massive funding landed in December. The company raised $155 million in the round led by Sandton Capital Partners, and landed with participation from Kesterl0X1, Blue Sky Capital Managers and Florence Capital.
“With a unique vertical integration approach, Nscale is building a hyperscale AI platform to power AI at scale,” Sandton co-founder Rael Nurick said in a press release.
Nscale said at the time it aimed to grow its AI data center pipeline from 300 megawatts the previous year to 1.3 gigawatts, running 350,000 GPUs by the end of 2027.
In comparison, CoreWeave said at a bank meeting last week that its portfolio consists of “around 2.2 gigawatts of capacity that will become online.” The company said in its IPO prospectus in March that 32 data centers operate 250,000 GPUs.
For Payne, the founder of Nscale, it was a whirlwind for several years. He was executive chairman of Arcon, but also served as chief of Australia’s Battery Future Acquisition Corp.
He does a lot of work in front of him.
Building an AI data center with expensive GPUs is a capital-intensive process that historically requires large amounts of debt. CoreWeave raised a total of $12.4 billion in debt by the end of 2024, in addition to raising more than $1 billion in equity funding prior to the IPO. It announced a $1.5 billion bond sales in July after offering $2 billion in debt in May.
According to Bloomberg, Nscale was looking to raise $1.8 billion through a private credit agreement led by a Goldman Sachs banker earlier this year.
In a December video related to Nscale’s equity funding, Payne called it “one of the biggest series in European history growing up in the UK.” He said the company will deploy up to 4,000 GPUs in data centers in Norway and will use cash to develop capacity of up to 180 megawatts in its portfolio.
The goal was to deploy 50,000 GPUs by the end of 2025 and 150,000 by the end of next year.
“The key challenge we see in the market is a significant increase in density at the GPU level,” he said. “This funding allows us to scale up materially,” he said, “becoming one of Europe’s biggest players.”
– CNBC’s Arjun Kharpal contributed to this report.
Watch: Openai Reorganization Clears the Hurdles
