TradingKey - Nvidia (NVDA)-backed UK AI cloud provider Nscale has officially launched its US listing process, seeking to raise more capital for its global data center expansion while the market continues to favor artificial intelligence infrastructure.

On September 18, Nscale submitted an S-1 registration statement to the US Securities and Exchange Commission and applied to list on the New York Stock Exchange under the ticker symbol "NSCL." The offering size and price range have not yet been determined. Goldman Sachs (GS), JPMorgan (JPM), and Morgan Stanley (MS) will serve as lead underwriters. According to Nscale's listing announcement, the registration statement has not yet become effective, and the final listing timing remains subject to regulatory review and market conditions.

According to the Financial Times citing people familiar with the matter, Nscale could seek a valuation of up to $35 billion in this IPO. The company also plans to open subscription channels to UK individual investors through RetailBook, allowing local retail investors to participate in this US listing.

What Is Nscale?

Nscale is an AI infrastructure company headquartered in London, UK, with operations covering data centers, power, GPU compute power, and cloud software. Unlike platforms that primarily provide traditional cloud computing services to enterprises, Nscale focuses on serving AI model training and inference demands by constructing or leasing data centers, deploying large-scale GPU clusters, and then leasing compute capacity on a long-term basis to artificial intelligence companies, technology firms, and government clients.

The company was spun off in 2024 from Australian cryptocurrency infrastructure firm Arkon Energy, after which it rapidly shifted its business focus from Bitcoin mining to AI compute power. Nscale currently has about 450 employees in the UK and has become one of Europe's larger AI infrastructure operators in terms of funding raised.

As of the end of August, the compute capacity in operation at the company's self-owned data centers was about 7 megawatts, with another 48 megawatts operating in leased facilities, representing contract values of approximately $2.6 billion. In addition, Nscale has about 1.3 gigawatts of projects in the planning and construction stages, and its future revenue growth largely depends on whether this capacity can be powered on and delivered to customers on schedule.

In March this year, Nscale completed a $2 billion funding round at a valuation of around $14.6 billion, with investors including Nvidia, Dell (DELL) and Nokia (NOK). According to Dealroom data, the company has accumulated about $3.7 billion in equity financing while raising over $5 billion in debt capital. If its IPO valuation reaches $35 billion, it would mean Nscale's market value will have more than doubled in about six months.

Nvidia Acts as Supplier, Investor, and Customer Simultaneously

The partnership between Nscale and Nvidia is not limited to GPU procurement. Prospectus filings show that Nvidia is both Nscale's core chip supplier and an investor, computing customer, and credit support provider for certain projects.

Nvidia has invested more than $2 billion in Nscale and signed computing lease agreements worth approximately $1.2 billion. In addition, Nvidia guarantees $860 million in lease obligations assumed by Nscale for a facility in Texas. While this relationship helps Nscale obtain GPUs, capital, and long-term orders, it also means the company has a high degree of reliance on the Nvidia ecosystem.

Both parties are also involved in UK AI infrastructure projects. Nscale plans to collaborate with Nvidia and OpenAI to expand local UK computing power, and work with Microsoft (MSFT) to build a large-scale AI supercomputer in Loughton, England. Nvidia previously stated that it would invest £500 million in Nscale to support a deployment plan for up to 300,000 GPUs.

For Nscale, these collaborations not only provide highly predictable customer demand, but also help build credibility when raising capital and securing data center resources. For Nvidia, investing in Nscale can encourage more cloud service providers to adopt its chips and extend GPU demand into the UK and European markets.

Massive Orders Coexist With Capital Pressure

Nscale's most high-profile project is located in West Virginia, where the company has signed a six-year compute leasing agreement valued at up to $45 billion with Anthropic, planning to serve it through a flagship data center currently under construction. The campus is expected to begin phased operations by the end of 2027, accompanied by the construction of a 2-gigawatt natural gas power facility, with total construction costs potentially reaching around $70 billion.

This contract provides Nscale with substantial long-term revenue visibility, but it also underscores the capital-intensive nature of its business model. Data centers require massive upfront capital to build out power, cooling, and computing infrastructure, while revenue is typically recognized gradually only after capacity is delivered. Should construction face delays, power supplies fall short, or customers adjust their compute demand, the company could simultaneously face rising costs and deferred revenue.

Financial data already reflects this pressure. According to the Financial Times, Nscale generated revenue of approximately $141 million in the first half of this year, with a net loss approaching $1 billion, driven in part by non-cash items. Relative to the $35 billion valuation the company is seeking, its current revenue scale remains small, meaning investors are effectively pricing in compute capacity that has yet to come online, long-term contracts, and future expansion capability.

Meanwhile, the timing of an IPO is also subject to uncertainty. Anthropic CEO Dario Amodei recently called on the industry to proceed with greater caution in advancing frontier AI model development, while OpenAI CEO Sam Altman and Elon Musk have also voiced concern over AI safety risks. These comments briefly triggered declines in AI chip and data center stocks, as the market worried that a slowdown in tech development could weaken future capital expenditure demand.

However, even if the pace of frontier model development is adjusted, commercialized AI products still require substantial inference compute capacity. Whether Nscale wins investor approval ultimately depends on whether the company can deliver its 1.3 gigawatts of capacity currently under construction on schedule, translate its partnerships with Nvidia, Microsoft, OpenAI, and Anthropic into sustained cash flow, and manage the debt and financing risks brought by rapid expansion.

Find out more