Nscale Secures Nearly $3 Billion for U.S. Data Center Campuses

On August 31, Nscale, a cloud infrastructure provider, said it had locked in financing commitments of roughly $3 billion to build two data center campuses in the United States. The money comes in the form of two senior secured delayed-draw term loans, one supporting a campus in Ward County, Texas, with a facility of up to $1.85 billion, and a second for a site in Madison County, North Carolina, with capacity of up to $1.2 billion, according to the company.

The announcement is a measure of how quickly money is moving into AI infrastructure. Nscale is a relatively young company in a crowded field of cloud providers, but it has found a niche supplying compute capacity to AI developers who need more power than public clouds can rent them. The company has said it aims to build out a national network of data centers, and the two new campuses are central to that plan. Ward County sits in West Texas, where land and power are comparatively cheap and where several AI data center projects have clustered. Madison County, in western North Carolina, gives the company a foothold in the Southeast, near the region’s growing tech corridor.

The financing structure reflects the realities of the data center business. Delayed-draw term loans allow a borrower to draw funds as construction progresses rather than paying interest on the full amount immediately, a feature well suited to projects that take years to complete. Senior secured lenders sit at the top of the capital stack, ahead of equity holders, a position that reflects both the size of the commitment and the caution of lenders in a sector where projects sometimes stall. People familiar with the deals said the terms were set after extensive diligence on power availability and tenant demand, the two risks that most often sink data center projects.

The Texas and North Carolina sites are part of a broader wave of construction driven by the AI boom. Data center capacity in the United States is being built faster than at any point in the industry’s history, and the companies doing the building range from hyperscalers like Amazon and Microsoft to specialist providers like Nscale. The math of the boom is simple: AI models require enormous computing power, and every large model training run consumes energy on the scale of a small city. That demand has made data centers one of the most active corners of commercial real estate and infrastructure finance.

The announcement also shows how far capital markets have moved to finance the build-out. Two years ago, lenders were wary of financing speculative data center capacity. Today, banks and credit funds are competing to lend, and terms have loosened as confidence in tenant demand has grown. Nscale’s financing, spread across two separate facilities, suggests lenders are willing to commit large sums in stages, tying each tranche to a specific campus rather than to the company as a whole.

For the counties involved, the projects bring jobs and tax revenue, along with the questions that accompany large data centers: power consumption, water use and the strain on local grids. Ward County has seen several similar announcements, and local officials have generally welcomed the investment. Madison County is newer to the data center economy, and the project there will test how quickly rural communities adapt to industrial-scale computing infrastructure.

Nscale declined to disclose who its tenants will be, and the company’s customer base has been a subject of speculation in the industry. Analysts said the financing would help the company compete for large AI contracts, but cautioned that the sector is crowded and that demand for data center capacity, while strong today, could cool if AI investment slows. “The lenders have priced in a lot of growth,” one infrastructure analyst said. “The question is whether the tenants show up at the pace the construction schedule assumes.”
The timing of the announcement is no accident. Data center financing is flowing at record levels, and lenders are choosing projects more carefully than they did in the early days of the AI boom, when any site with a power agreement could find funding. Nscale’s two-campus approach, with separate facilities for each project, signals to the market that the company is building for the long term rather than chasing a single headline deal. It also spreads execution risk: if one campus slips, the other can proceed without dragging the whole financing package down.

For now, the company is moving forward. Groundwork at the two sites is expected to begin within months, and Nscale has said it plans to bring the campuses online in phases, starting in 2027. The commitments announced Monday give it the funding to proceed, and the two projects will test whether a mid-sized cloud provider can deliver on the scale its lenders have underwritten.

Related Posts

  • September 6, 2026
  • 6 views
Anthropic Moves Its IPO Filing to Late September

The bankers and lawyers running Anthropic’s initial public offering had told investors to expect the company’s registration documents as soon as this week. The calendar has moved. Anthropic now plans…

  • September 6, 2026
  • 6 views
OpenAI Quietly Revises GPT-6 Astra Scores After Launch

When OpenAI released GPT-6 Astra on Sept. 3, the launch post carried the usual furniture of a modern model debut: coding results, speed comparisons and a figure for how often…