Nvidia is working with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR on financing platforms intended to mobilize more than $500 billion for AI infrastructure. As part of the plan, Nvidia may also provide limited guarantees on the future value of its GPUs, helping lenders treat AI hardware as collateral for long-term financing.
The six financial groups will establish independent pools of capital that can finance AI data centers and computing infrastructure. Nvidia said the structure is intended to bring more institutional capital into AI infrastructure without requiring Nvidia itself to fund most of the construction.
In its official announcement, Nvidia said the financing would support hardware sales, software adoption, and long-duration contracts for AI computing capacity. The agreements remain subject to final terms.
Nvidia Wants Aging GPUs to Retain More Value
A key part of the financing model is Nvidia’s effort to support the residual value of GPUs after they have spent several years inside a data center. Financial Times reporting indicates Nvidia could absorb losses if certain chips used as collateral fall below agreed valuation levels, with support potentially covering up to 25% of some financed projects.
That could make it easier for lenders to finance AI infrastructure because older GPUs would have a clearer expected resale or leasing value. Nvidia CEO Jensen Huang argued in an Nvidia blog post that AI computing infrastructure can be moved between customers, cloud providers, and operators as demand changes.
The model also creates risk for Nvidia. If demand for AI computing weakens and used GPU prices decline, the company’s guarantee obligations could rise at the same time that weaker demand affects its hardware sales.
Private Capital Could Finance More AI Data Centers
Huang has rejected comparisons with vendor financing models in which technology suppliers directly lend customers money to buy their products. Nvidia says the new initiative instead relies primarily on independent capital supplied by large asset managers, while Nvidia provides limited support around the value of its computing equipment.
The approach could create a larger secondary market for Nvidia hardware as newer GPU generations arrive. Older processors could continue serving customers with less demanding workloads rather than losing most of their economic value when newer models enter the market.
Nvidia describes the broader concept as turning AI compute into an investable infrastructure asset rather than treating servers like conventional technology equipment that quickly depreciates. The company argues that continued demand across multiple customers and operators can support that residual value over longer periods.
Featured image credits: EdTech Stanford University School of Medicine via Flickr
For more stories like it, click the +Follow button at the top of this page to follow us.