Nvidia weighs playing the cosigner for Nscale’s AI data center in West Virginia
A debt guarantee means Nvidia promises to support Nscale’s borrowing if Nscale cannot meet certain obligations. Nvidia would not necessarily pay immediately. Instead, its backing would reduce lenders’ risk and make financing for the project easier to secure. The proposed deal is called a residual value guarantee. If Nscale defaulted on construction payments, Nvidia would temporarily take over the facility, help find a buyer, and help find a new tenant. That goes beyond Nvidia’s earlier support for cloud providers, which mostly covered leases and chip purchases. The arrangement would deepen Nvidia’s financial ties to Nscale. It could also help protect Nvidia’s chip sales because Paradox is expected to use Nvidia’s next-generation Vera Rubin chips at the site. The risk is significant because Nscale is pursuing ambitious projects while seeking a public valuation of up to $35 billion.
What does it mean for Nvidia to guarantee Nscale’s debt, or act as a co-signer on its loans?
A debt guarantee means Nvidia promises to support Nscale’s borrowing if Nscale cannot meet certain obligations. Nvidia would not necessarily pay immediately. Instead, its backing would reduce lenders’ risk and make financing for the project easier to secure.
The proposed deal is called a residual value guarantee. If Nscale defaulted on construction payments, Nvidia would temporarily take over the facility, help find a buyer, and help find a new tenant. That goes beyond Nvidia’s earlier support for cloud providers, which mostly covered leases and chip purchases.
The arrangement would deepen Nvidia’s financial ties to Nscale. It could also help protect Nvidia’s chip sales because Paradox is expected to use Nvidia’s next-generation Vera Rubin chips at the site. The risk is significant because Nscale is pursuing ambitious projects while seeking a public valuation of up to $35 billion.
What exactly is Nscale trying to finance in West Virginia, and what role would Paradox play in the project?
Nscale is seeking financing for the electricity and construction costs of its West Virginia data center campus. The immediate project involves building large computing facilities that can host advanced AI workloads. Securing those facilities requires substantial capital before they can generate revenue.
Paradox has signed a $44.6 billion contract covering the campus’s first building. That building is planned to have 460 megawatts of capacity and is expected to use Nvidia’s next-generation Vera Rubin chips. The companies are also discussing a second building deal worth roughly $45 billion.
Paradox would therefore be a major customer and a source of demand for the campus. Its involvement likely helps explain Nvidia’s interest in keeping the project on track. Without a large customer, lenders and equipment suppliers could face more uncertainty about whether the facility will earn enough to support its financing.
How large is the planned project, in terms of construction cost, electricity capacity, chips, and potential company valuation?
The West Virginia project is large in several different ways. Paradox’s contract for the first building is valued at $44.6 billion. That building is planned to provide 460 megawatts of electricity capacity, showing how much power the intended AI computing operation would require.
Nscale had about 25,000 advanced AI chips running in August. It plans to deploy nearly 500,000 chips, so its current operating base is only a fraction of its target. The first building is expected to use Nvidia’s next-generation Vera Rubin chips, with a second building under discussion at roughly $45 billion.
Nscale is also seeking to go public at a valuation of up to $35 billion. These figures show both the opportunity and the pressure. The company must raise money, build infrastructure, secure power, and deploy chips while investors question whether its ambitious projects can be delivered.
What would happen if Nscale could not make its construction payments under the proposed residual value guarantee?
The proposed residual value guarantee is designed to protect the facility’s value if Nscale cannot keep making construction payments. Nvidia would step in under the agreed conditions, giving lenders a stronger fallback than Nscale’s promise alone.
The key mechanism is temporary control. If Nscale defaulted, Nvidia would take over the West Virginia facility for a period. It would then help locate a buyer for the property and a new tenant to use it. This could help lenders recover value from an unfinished or operating data center.
That outcome would still expose Nvidia to financial and operational risk. The guarantee would tie Nvidia more closely to Nscale’s ability to complete the campus and attract customers. It also shows how the company is supporting AI infrastructure beyond selling processors, particularly when ambitious data center projects face financing and construction problems.
Why would Nvidia take on financial risk for Nscale instead of simply selling it chips?
Nvidia would take this risk because Nscale’s success could create a large, dependable customer for its AI chips. Paradox is one of the largest buyers of AI computing power, and its West Virginia project is expected to use Nvidia’s next-generation Vera Rubin chips.
The guarantee could help Nscale obtain financing for electricity and construction. In return, Nvidia would have a stronger chance of seeing the campus built and equipped with its hardware. The proposed arrangement is more extensive than Nvidia’s earlier support, which mostly involved leases and chip purchases.
Nvidia has already committed about $1 billion in convertible financing to Nscale and has extended roughly $170 billion in investments and credit support to neoclouds and large customers. The strategy deepens relationships with infrastructure providers. It also creates exposure if projects fail, power is delayed, customers change suppliers, or Nscale cannot deliver its plans.
Why do AI data centers need so much electricity and specialized hardware, and how can shortages of either one delay or cancel a project?
AI data centers require large amounts of electricity because specialized processors perform vast numbers of calculations for training and running AI systems. The facilities also need many of those processors, plus cooling and other supporting equipment. Their scale makes power and hardware central to whether a project can operate.
A project can be delayed when the electricity grid cannot supply enough power on schedule. Microsoft pulled out of a planned Nscale site in England after grid power was pushed into the 2030s. Hardware constraints matter too: Nscale had about 25,000 advanced AI chips running in August but planned nearly 500,000.
These shortages can undermine financing and customer commitments. A building without power cannot deliver computing capacity, while a building without chips cannot use its electrical capacity. Nscale’s West Virginia campus also faced construction trouble, including runoff that flooded nearby homes, adding to the project’s execution challenges.
What are neoclouds, how do they make money by renting AI computing power, and why are investors worried about their debt and ability to deliver?
Neoclouds are companies that rent out AI computing power instead of relying on a cash-rich core business. They build or obtain data center capacity, equip it with advanced chips, and charge customers for access to that computing resource. Their business depends on turning expensive infrastructure into steady rental revenue.
The model requires major spending before revenue arrives. A neocloud may need to finance buildings, electricity, chips, and operations while waiting for customers to use the capacity. If construction or power is delayed, the company may still face debt payments without the expected income.
Investors have become concerned about that risk. Shares of CoreWeave, Nebius, and IREN had each fallen more than 30% in October as of the prior week, CNBC reported. Nscale also had only about 25,000 advanced AI chips running against a plan for nearly 500,000, raising delivery concerns as it seeks a public valuation of up to $35 billion.
This brief was written by AI from the original reporting and checked by other models. Names, figures and quotes come from the source; read it for full context.
Read more in the JupiteX app
Pulse is free. New stories every 4 hours, each one broken into the questions that explain it.
Or read more news on the web