NVIDIA has introduced a revenue-sharing GPU supply model, providing token credits to AI startups wit...
The AMW Read
A structural shift in how GPU compute is financed and distributed β NVIDIA is internalizing the capital-cycle risk that previously constrained AI infrastructure growth, resolving an open debate about whether compute access or capital was the binding constraint for startups.
NoveltySignificance
AI Infra Β· Player Map
NVIDIA has introduced a revenue-sharing GPU supply model, providing token credits to AI startups with limited capital upfront. The program, announced by CFO Colette Kress, connects AI cloud operators with startups, offering GPU credits in exchange for a share of future sales revenue. NVIDIA also guarantees compute capacity for underutilized GPUs, effectively acting as a backstop for cloud providers. Initial partners include Australia's Sharon AI and Singapore's Permus Technology, with deployments involving up to 40,000 Grace Blackwell GB300 GPUs in a planned Indonesian AI factory.
#NVIDIA#revenue-sharing GPU model#AI infrastructure funding#GPU token credits
