Nvidia has paused some deals in a financing program that offered credit support to AI cloud companies in exchange for a share of their revenue, the Wall Street Journal reported Thursday, citing people familiar with the matter. Reuters corroborated the Journal’s account.
The chipmaker stepped back last week, the Journal said, less than two months after it announced the initiative. Reuters said the program was designed to help small AI cloud firms finance purchases of Nvidia processors. Nvidia had offered to rent unused capacity back if those firms could not sell it, giving them a guaranteed buyer and a cleaner path to borrow money. Nvidia would then make money twice: once on the hardware sale, and again as a cut of the cloud revenue those chips produced. Under the proposed terms, the Journal reported, Nvidia would take 50 percent of revenue above a set threshold.
The Journal named the effort as part of Nvidia’s AI Compute Partnership. It is not a cancellation. The paper said Nvidia may revamp the program or fold it into another one. An Nvidia spokesperson told Reuters: “The new business model … that opens up compute access to the fast-growing AI ecosystem is still in place and continues to evolve due to high demand.”
The pause comes as investors have grown louder about circular AI deals, Reuters reported. Nvidia said on this week’s earnings call that the model could drive billions of dollars in revenue over the medium to long term. The company this month also helped arrange $500 billion in financing from major U.S. banks for customers, and agreed to guarantee up to $105 billion to help OpenAI lease a data center, Reuters said.
Control was the other problem. The Journal reported that some Nvidia employees flagged antitrust concerns to current and potential customers, and that there were sensitivities about how far Nvidia can tell those customers how to run their business. In the program’s early weeks, Nvidia irked some potential partners. It told some providers they could rent chips only to approved customers, and it preferred that capacity be spread across smaller firms rather than concentrated with one large renter.
This is a pause on the revenue-share layer, not a pause on selling GPUs. It is about who gets to rent the chips, and who gets a cut.
Nvidia still wants the ecosystem financed. It is less eager, for now, to write the rental rules.