TL;DR — Key Takeaways
NVIDIA Corp. has abruptly pulled back from a controversial financing program that offered credit backing to artificial intelligence (AI) cloud providers in exchange for a share of their revenues, following internal concerns over potential antitrust exposure and friction with prospective partners.
Introduced in July as the AI Compute Partnership, the initiative was intended to help smaller cloud firms secure loans for costly NVIDIA hardware. Under the plan, the semiconductor giant promised to rent unused graphics processing unit (GPU) capacity itself if a provider failed to find third-party buyers. In return, NVIDIA stood to profit twice: First from the initial hardware sale, and subsequently through a 50% cut of rental revenues earned above an agreed cost threshold.
However, the chipmaker halted the program just weeks after its debut, according to report details first published by The Wall Street Journal. The move came after NVIDIA employees expressed concerns to clients that the arrangement’s tight terms could draw regulatory scrutiny.
During early operations, NVIDIA insisted on vetting its partners’ end-customers and mandated that cloud capacity be distributed across multiple smaller AI firms rather than concentrated with single large buyers. Several cloud providers objected to the terms, maintaining that client selection should remain their own decision.
Initial partners named under the paused partnership included Sharon AI and Firmus Technologies.
Despite the setback, NVIDIA signaled that its broader strategic vision remains active.
“The new business model we introduced in July that opens up compute access to the fast-growing AI ecosystem is still in place and continues to evolve due to high demand,” an NVIDIA spokesperson said. Sources familiar with the matter noted the company could eventually revamp the arrangement or fold it into another effort.
The scale of the venture was revealed in a recent quarterly filing, which detailed $36 billion in total commitments tied to agreements spanning a typical six-year duration. Chief Financial Officer Colette Kress addressed the revenue-sharing model during an investor call, noting that while the new income stream had the potential to generate billions over time, the company’s financial risk remained low because hardware could readily be redeployed. Kress also pushed back against critics who characterized the setup as “circular financing.”
The initiative forms part of NVIDIA’s aggressive strategy to support its customer ecosystem, alongside direct investments in AI research labs and partnerships with major financial institutions — including BlackRock, Blackstone, and Apollo Global Management — to mobilize over $500 billion in infrastructure capital.
The capital deployment strategies have drawn sharp commentary from high-profile market observers, including “The Big Short” investor Michael Burry, who recently dismissed the financing efforts as a “Wall Street stunt.”
Stephanie Walter, practice lead for AI Stack and Enterprise Application Development at HyperFRAME Research, said under the reported structure, NVIDIA could “simultaneously be the GPU supplier, credit backer, buyer of unsold capacity, revenue-sharing partner, and an influence over which customers receive that capacity.”
“That makes it harder to separate independent market demand from demand supported by NVIDIA’s own commitments, while giving the dominant AI chip provider considerable influence over its customers’ commercial decisions,” she said. “Pausing the program appears prudent. I hope NVIDIA can preserve the access benefits while creating clearer boundaries and greater independence for participating cloud providers.”
The structural pivot comes amid stellar financial performance for the tech giant. NVIDIA this week posted record second-quarter revenue of $96.22 billion, easily topping Wall Street expectations of $92.11 billion, while forecasting $108 billion for the third quarter.
