Nvidia Launches Revenue-Sharing Model for AI Cloud

Nvidia introduces a new revenue-sharing and credit-support model for AI cloud partners, starting with Sharon AI and Firmus Technologies.

Nvidia CFO Colette Kress co-authored a blog post detailing a new revenue-sharing and credit-support model for AI cloud partners. Under this structure, Nvidia collects hardware revenue plus a percentage of the revenue generated by that hardware, while developers receive token credits in exchange for future sales. The first partners, Australia’s Sharon AI and Singapore’s Firmus Technologies, will deploy up to 210,000 GPUs. Sharon AI’s six-year deal covers 72 MW of Australian data center capacity designed to Nvidia’s DSX AI factory standards, scaling to 40,000 Grace Blackwell GB300 GPUs. Firmus is building a 360 MW DSX-aligned campus in Batam, Indonesia, for 170,000 GPUs. Nvidia also recently participated in a $110 billion OpenAI funding round and backed xAI’s $20 billion Colossus 2 financing.