Nvidia has unveiled a new, optional financing model that lets AI startups get access to its sought-after GPUs now and pay later — partly by handing back a share of their future revenue.
According to Tom's Hardware, the arrangement works like a financing vehicle: instead of only selling chips outright, Nvidia can take a cut of the cloud revenue those same chips help generate, effectively earning twice on the same silicon. The Next Web describes it as a "compute now, payment later" offer, paired with credit support, that opens Nvidia's hardware to startups that previously couldn't afford it.
HotHardware reports the revenue-sharing and credit-support model is rolling out with AI cloud partners including Sharon AI and Firmus, tying GPU sales to those partners' long-term cloud earnings. The Edge Singapore says the goal is to build business among researchers and young companies that lack the capital for large-scale AI resources.
The Decoder frames the strategy more broadly: by bankrolling smaller players, Nvidia is acting almost like a central bank for AI startups and working to loosen Big Tech's grip on its chip business, since a handful of giant cloud customers currently dominate demand.
Investors were not immediately impressed. Barron's and MSN both note that despite the plan to expand Nvidia's customer base, its stock was falling, with the shares lagging the broader semiconductor sector this year.
Why it matters: the move could put cutting-edge AI computing within reach of far more startups, while deepening their financial ties to the single company that already dominates the market for AI chips.