British cloud computing company Nscale, which is backed by Nvidia, has announced a steep rise in revenue as it looks to go public in a planned initial public offering in the United States in the weeks ahead, providing investors with a more detailed look at the performance of one of the companies that could benefit from the rapid growth of artificial intelligence computing. Revenue in the first half of 2026 hit $140.6 million, an increase of 1,252% compared to $10.4 million during the same period last year.
As AI models evolve and become more complex, the surge in revenue is expected to stick around for the foreseeable future. But Nscale’s financials also underscore the high expense of setting up the infrastructure needed to power the AI industry. The company earned a net loss of $1.02 billion in the first half of this year, compared with a net loss of $368.9 million in the same period last year.
As investors continue to seek deals in the AI space, but have grown more choosy with their spending and valuations, Nscale is planning to go public at a time when there is still investor interest. The firm is reportedly aiming for listing to be valued at more than a multibillion dollar, and could see a valuation of some $30 billion. The valuation would be a significant jump from the $14.6 billion that it was valued at in a funding round in March.
This possible listing will be a new benchmark for public market investors’ confidence in companies further down the AI infrastructure supply chain. The AI chipmakers and software developers have garnered a lot of interest, but the companies that supply the computing power, electricity, data centers and cloud systems that power AI models are also big winners in the AI technology boom.
“It’s not as euphoric as a couple of months ago, but the setup to do these deals is adequate,” said Matt Kennedy, senior strategist at Renaissance Capital, a firm that sells IPO-related research and ETFs.
“There seems to be more resilience on the “neocloud” product side vs. higher up the stack AI stocks,”
This separation is crucial as the demand for AI computing continues to rise amidst growing investor fears of over-investment and rapidly appreciating valuations. Other specialist cloud vendors have also seen evidence of ongoing demand. Nebius has been raising the price of its cloud-based AI services, while CoreWeave has been holding onto contracts at higher prices. The developments show that companies working on the development and operation of AI systems still need to deal with huge computing capacity.
Nscale aims to harness that demand, by pairing together several key components of the AI infrastructure ecosystem into a business. The company offers computing resources, power capacity, data centers and software for artificial intelligence training and operation. At present, it has a network of over 10 gigawatts of power pipelines and is present in 14 regions.
It has gone through phenomenal growth in contracted business. In about two and a half years, Nscale’s total contract value reached over $103 billion, from a mere $100 million. This is the value of agreements related to future business, and hence it is not the same as revenue already recognized in the company’s financial statements. Yet, the growth rate shows the magnitude of the pledges in the arena of AI computing infrastructure.
As AI developers race to access computing power, the company has made a number of big deals. One of the highlights was an agreement with Anthropic that is said to be worth $45 billion, in which Anthropic will provide West Virginia’s Nscale data center campus with AI cloud computing capacity. This kind of deal highlights the need for specialized cloud providers to find more ways to grow their computing infrastructure and power supply as AI workloads increase.
There are bad news stories in the headlines, but there are still billions that still come to this space. While there remain huge compute and power infrastructure challenges with AI, Nscale’s $100-plus billion-plus in total contract value was a big talking point, Kennedy said.
But the fast expansion carries some dangers that would need to be weighed by any would-be investor. Many of the customers represented a large portion of Nscale’s revenue. The top customer generated 52% of the revenue for the first-half of 2026. This focus can make them vulnerable if there is a significant customer order reduction or order delays and changes in their infrastructure plan.
Nscale’s prospectus also names Microsoft and Anthropic as two customers that are expected to be significant for the company’s business in the future. Major tech and AI firms may help keep the growth going, but the company’s future revenue trajectory will be based on how quickly these ties turn into reported revenue and long-term cash flow.
One of the other major considerations is that Nscale’s capital needs are very high. Cost of building AI data centers involves the purchase of specialized computing hardware, electricity infrastructure, land and facilities. The firm recently signed contracts to sell $3.1-billion in convertible bonds, among them $1-billion to Nvidia, the filing said. The funding will also serve as further capital as Nscale strives to expand its infrastructure, and as proof of the amount of capital needed to be competitive in the AI cloud space.
Nscale’s product is available to rival offerings such as CoreWeave, Nebius and Crusoe and Lambda, which are also vying to break into a fast-growing market. The competition will continue to be fierce as AI firms continue to increase model training, inference, and other computation-intensive tasks. Reliable power and data center capacity are especially critical with rapid demand growth as providers scramble to keep up.
