Nscale Files for $3 Billion IPO on New York Stock Exchange to Fund AI Computing Expansion
Key Takeaways
- •Nscale filed with the SEC on September 18 for a New York Stock Exchange IPO aiming to raise up to $3 billion and will trade under the ticker NSCL.
- •The company generated $140.6 million of revenue in the first half of 2026, up 1,252% year over year, but recorded a $1.02 billion net loss and holds more than $8 billion in total debt.
- •Nscale's remaining performance obligations total $56.4 billion, with its client roster including Microsoft, OpenAI, and Anthropic, whose multi-year contract alone is valued at $45 billion.
- •The company raised $3.1 billion in funding during 2026, with its $2 billion Series C round valuing it at $14.6 billion, and it is targeting up to 10 gigawatts of power capacity.
- •CoreWeave, which went public earlier in 2026, is pursuing the same AI compute market, offering investors an existing listed comparable for the sector.

Nscale, a London-based AI cloud infrastructure company founded just two years ago, has filed for an initial public offering on the New York Stock Exchange. The company is seeking to raise up to $3 billion to fund its rapidly expanding operations in the AI computing space.
The registration statement was filed with the U.S. Securities and Exchange Commission on September 18, placing Nscale in a growing queue of AI infrastructure firms attempting to tap public markets while investor appetite for anything GPU-adjacent remains voracious. Once listed, the company will trade under the ticker NSCL. The filing puts the company's detailed financials on public record and begins the standard path to listing, in which terms are refined and shares are priced closer to the debut.
Revenue Is Soaring, but So Are Losses
Nscale posted $140.6 million in revenue during the first half of 2026, representing a 1,252% increase year over year. Over the same six-month period, however, the company reported a net loss of $1.02 billion and carries more than $8 billion in total debt.
The figures underline the capital intensity of AI infrastructure: providers spend heavily upfront on datacenters and GPU capacity long before contracted revenue translates into profit, which is why a public listing would open a funding channel beyond private capital.
Its remaining performance obligations—essentially contracted future revenue—stand at $56.4 billion, a figure that reflects the scale of commitments the company has already secured from customers. Set against $140.6 million of first-half revenue, that backlog makes the pace at which contracts convert into recognized revenue a key indicator of how the buildout is progressing.
A Client List That Reads Like an AI All-Star Roster
Microsoft, OpenAI, and Anthropic all count among Nscale's clients. The Anthropic relationship alone is valued at $45 billion across a multi-year deal, making it one of the largest infrastructure contracts in the AI sector.
The company also maintains close ties with Nvidia, which supplies the GPU resources that form the backbone of Nscale's cloud computing platform.
Josh Payne founded Nscale in 2024, positioning the company to ride the wave of demand that emerged as large language models and AI applications began requiring unprecedented amounts of computing power.
From Startup to $14.6 Billion Valuation in Two Years
The company raised $3.1 billion in total funding during 2026 alone. Its Series C round, worth $2 billion, valued the business at $14.6 billion. A public listing would move the pricing of the company from private rounds to open markets for the first time.
Nscale is targeting capacity of up to 10 gigawatts of power. For context, a single gigawatt can power roughly 750,000 homes. Targets on that scale underscore how power access has become a central factor in AI datacenter expansion.
What the IPO Means for the AI Infrastructure Market
The case in favor is straightforward: AI computing demand is growing faster than supply, Nscale has locked in contracts worth tens of billions of dollars, and its client roster includes companies that are themselves valued at hundreds of billions of dollars. The concerns are equally clear: more than $8 billion in debt, a billion-dollar loss in just six months, and a business model that requires continued massive capital deployment before reaching profitability.
Nscale is not the only company racing to build AI compute capacity. Specialized providers like CoreWeave, which went public earlier in 2026, are chasing the same market, giving public-market participants an existing listed comparable for the sector.
For investors evaluating the offering, the $56.4 billion backlog will likely be the most scrutinized number in the prospectus. Between the filing and any listing, the watch items are the SEC's review of the registration statement, updated filings with final terms, and disclosures showing how contracted revenue converts as new capacity comes online.