Nvidia broadens “AI factory” push with revenue-sharing and credit-support deals for cloud partners

Nvidia is expanding the way it monetizes the buildout of “AI factories,” signing arrangements with AI cloud providers that include revenue sharing and credit support, according to a report citing the deals as part of the company’s push to scale AI computing capacity. The move highlights how demand for training and inference horsepower is increasingly tied not just to chip shipments, but also to who absorbs the financial and operational risk of building capacity.

The core idea is to move AI infrastructure closer to where compute is needed, while aligning incentives between Nvidia and the cloud operators that run models. Nvidia has described its “AI Cloud ecosystem” as a network of purpose-built clouds designed to handle the rising token demand behind popular AI applications, combining accelerated computing, high-speed networking, and AI software as an integrated stack.

In its own descriptions of the ecosystem, Nvidia has said AI clouds have been co-designed with its broader full-stack infrastructure to support workloads ranging from training and fine-tuning to inference, agentic AI, and “physical AI” and “sovereign AI” deployments. The company has emphasized that configurations vary by partner and workload, but the theme is consistent: the hardware, networking, and software are meant to work together to deliver better economics such as “lowest token cost” and higher throughput per watt.

The reported revenue-sharing and credit-support elements add another layer. Revenue-sharing can tie returns to the actual performance and usage of the AI services running on Nvidia hardware, rather than only to upfront equipment sales. Credit support, meanwhile, can make it easier for cloud operators to finance new buildouts or expansion, potentially accelerating the timeline for new capacity.

Nvidia’s partnership framing also positions its ecosystem as global and regional, including support for “nearly every geography,” including “sovereign AI capacity” for frontier model builders, enterprises, startups, software providers, and national AI programs. Nvidia is effectively presenting the AI clouds as a distribution channel for its platform, while also positioning itself as a central infrastructure supplier for the next generation of AI workloads.

Jensen Huang, Nvidia’s founder and CEO, has said that “every company and every country needs AI factory infrastructure to turn data into intelligence,” arguing that Nvidia AI Clouds bring full-stack AI factories closer to regions and industries building new AI, from model training to real-time inference and AI agents. The reported deal structure described in the Yahoo Finance report is consistent with that broader narrative, where Nvidia’s role extends beyond selling GPUs into helping scale the capacity that uses them.

Still, key details remain undisclosed in the publicly available coverage used for this report. The Yahoo Finance item does not specify deal sizes, the duration of the agreements, which specific cloud companies are involved, or how revenue-sharing formulas and credit-support terms are calculated. Until more detailed disclosures emerge, investors and customers will have to judge the impact mainly through Nvidia’s broader AI cloud strategy and subsequent financial reporting, rather than through quantified guidance about these particular contracts.