British AI infrastructure provider Nscale announced on Friday that it has closed a $3.36 billion convertible-note financing round, a move that comes as the firm prepares to list on the New York Stock Exchange later this year. The funding, structured as a convertible note, highlights the enormous capital outlays required to construct large-scale AI data centers.
Lead investor hedge-fund Third Point contributed the bulk of the round, providing $2.36 billion that Nscale can draw immediately. An additional $1 billion will be supplied by existing backer Nvidia in mid-November, with both tranches set to convert into equity when the IPO closes. The convertible structure allows the investors to become shareholders after the public offering, aligning their interests with the company’s future performance.
Nscale filed its registration statement last week and, according to the Financial Times, expects a market valuation of roughly $35 billion on the NYSE. Bloomberg reports that the company aims to raise $3 billion through the offering, positioning the float among the largest technology listings of the year.
The firm originated two years ago when it was spun out of Australian cryptocurrency-mining group Arkon Energy. Its IPO filing indicates that Nscale has already secured more than $103 billion in contracts, reflecting strong demand for its AI-focused data-center services. These agreements span multiple sectors, including cloud providers, enterprise customers, and research institutions seeking high-performance compute capacity.
Nscale is actively constructing several large-scale data-center campuses, with projects announced in Norway and in West Virginia, United States. The sites are intended to host dense arrays of AI accelerators, positioning the company to meet the growing computational needs of generative-AI workloads. Both locations benefit from favorable power pricing and proximity to renewable energy sources, factors that are critical for the energy-intensive nature of AI training.
The size of Nscale’s financing round underscores the escalating capital intensity of AI infrastructure, where building and operating specialized data centers can require billions of dollars before revenue materializes. Analysts see the convertible-note approach as a way for investors to share upside while mitigating early-stage risk, a model that may become more common as the sector matures.