Loading page…
The proposed program paired capital for Nvidia hardware with a claim on cloud revenue, putting new attention on how far a chip supplier can reach into its customers’ businesses.
Listen to this story
Nvidia has reportedly put parts of its AI Compute Partnership Program on hold, potentially disrupting a financing model aimed at smaller cloud providers building Nvidia-based data centers. The proposal went beyond chip sales: Nvidia could fund hardware, take up to 50% of revenue above a threshold, and lease unused capacity. Employee antitrust concerns and partner objections to rental restrictions surfaced, though neither is confirmed as the cause.
The program launched less than two months ago and may be restructured or absorbed into other Nvidia initiatives.
Sharon AI and Firmus Technologies were identified as early participants in the proposed partnership.
Nvidia’s broader infrastructure-finance push reportedly includes a $500 billion capital initiative and up to $105 billion in guarantees tied to OpenAI leasing plans.
AI cloud providers seeking capital for Nvidia-heavy data centers now face uncertainty over one proposed route to funding. Nvidia reportedly suspended parts of its AI Compute Partnership Program less than two months after its introduction; the program offered financial support for Nvidia hardware in exchange for a share of the computing revenue it produced.
The precise reason for the reported suspension remains unclear. The initiative could be restructured or folded into other programs, while Nvidia says its broader business model remains active and is changing as demand for AI computing grows.
The structure addressed a difficult funding problem. Smaller AI cloud companies can need to commit billions of dollars to chips and data centers before enough customer contracts exist for conventional financing. Nvidia’s support could reduce that risk, but it also made the company a chip supplier, financier and potential participant in its customers’ revenue.
Under the reported initial proposal, Nvidia could receive half of a cloud provider’s revenue above a specified threshold.
Some Nvidia employees raised concerns that the arrangement could invite antitrust scrutiny, according to BigGo Finance. Potential partners also objected to proposed restrictions on how they could rent computing capacity. Those objections describe friction around the terms, not a confirmed explanation for the suspension.
The episode lands amid Nvidia’s wider push into infrastructure finance. BigGo Finance said Nvidia recently joined an initiative targeting more than $500 billion in third-party capital for AI infrastructure, and that the company offered guarantees of as much as $105 billion connected with OpenAI data-center leasing plans.
That wider role has prompted investor and analyst questions about circular financing: Nvidia can sell chips, help customers finance them and potentially share in the revenue they generate. The unresolved issue is whether the company can offer the capital cloud builders need while keeping those roles acceptably separate.
Editorial analysis
This is a stress test for Nvidia’s financing strategy, not necessarily a retreat from it. The proposed terms combined three forms of influence: support for hardware purchases, a claim on revenue above a threshold, and a potential lease backstop for unused capacity. That is more involved than simply helping a customer fund equipment. The signal to watch is whether any reworked program separates financing from revenue sharing and rental restrictions. Nvidia’s wider reported infrastructure-finance commitments make that distinction increasingly important for cloud operators and capital providers.
Loading discussion...
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.