Nscale Raises $3.36 Billion in Pre-IPO Financing

AI cloud infrastructure company Nscale Limited has raised $3.36 billion in pre-IPO convertible financing, one of the largest private capital raises in the sector as it builds data centers, power infrastructure and GPU clusters ahead of a planned public offering.

Third Point led the round. NVIDIA, Apollo, Citadel, Hudson Bay Capital, Abu Dhabi Investment Council and 8090 Industries also participated. Nscale said it has more than $103 billion in total contracted value across its business.

The deal comes in two parts. An initial $2.36 billion tranche funded at closing. NVIDIA accounts for the remaining $1 billion, which is expected to fund in mid-November 2026.

The convertible loan notes will automatically convert into ordinary shares when Nscale completes an initial public offering. NVIDIA's portion converts into non-voting shares. The structure ties investor returns directly to the company's move into public markets, though Nscale did not provide a timetable for its IPO.

"This marks a milestone for Nscale as we continue scaling our full-stack AI infrastructure to meet unprecedented global demand," founder and CEO Josh Payne said. "With the backing of these world-class investors, we are strongly positioned to accelerate our data center buildouts globally."

The size of the raise underscores the capital intensity of building AI computing infrastructure. Nscale operates a vertically integrated model that spans power generation, liquid-cooled AI data centers, GPU computing clusters and the software customers use to run AI training and inference workloads.

That integration reaches as far upstream as behind-the-meter power plants, which supply electricity directly to computing facilities. At the data center level, Nscale is developing liquid-cooled facilities built for the dense GPU systems that large-scale AI workloads demand. The company pairs that physical infrastructure with a cloud platform giving customers compute access for model training and inference.

Nscale serves hyperscalers, frontier AI model developers, AI-native companies and enterprises deploying AI applications. Its strategy is to control more of the infrastructure stack rather than relying entirely on third-party data centers or power arrangements, coordinating power availability, data center capacity, computing hardware and cloud software as deployments grow.

The company said it will deploy the new capital to accelerate development of that infrastructure globally, responding to rising demand for computing capacity from AI developers and enterprise customers.

NVIDIA's role stands out. The chipmaker committed $1 billion of the total round, deepening its financial stake in the AI infrastructure buildout that drives demand for its processors. Its investment is expected to close separately in November, subject to the stated funding timetable.

The broader investor group includes Davidson Kempner Capital Management, Qube Research & Technologies, Context Capital Management, Longaeva Partners, Wellington Management, Castleknight, Ghisallo Capital Management, LionTree Investment Fund, Javelin Venture Partners and Irving Investors.

The financing lands as access to power and data center capacity has become the central constraint on AI computing expansion. Building large GPU clusters requires substantial spending not only on processors and networking equipment but also on electrical infrastructure, cooling systems and facilities capable of handling increasingly power-dense hardware.

Nscale is betting that a single platform spanning software, compute and power can solve those bottlenecks. The new financing, according to the company, will support further expansion across each of those layers.

Goldman Sachs & Co. LLC acted as placement agent for the $3.36 billion raise.