Jensen Huang: Nvidia confident OpenAI keeps using its tech

Jensen Huang said Nvidia remains confident OpenAI will continue using its technology despite OpenAI's Jalapeño custom chip, arguing that Nvidia serves the full AI lifecycle rather than a single inference workload.

The Details: • OpenAI relationship: Huang said he has "100% confidence" Nvidia's technology and economics will remain compelling for OpenAI, which he expects to remain a customer and partner for a long time. • Platform differentiation: He contrasted Nvidia's broad AI-factory platform—covering data processing, training, post-training, and agentic inference across clouds—with ASICs designed for a specific cloud, service, or inference task. • Global deployment: Huang argued that AI services will expand globally and that their future data centers will not necessarily be owned or built solely by the AI companies themselves, supporting demand for portable, broadly deployable Nvidia infrastructure. • Unconstrained demand: He said unconstrained revenue would be "a lot higher," calling the revenue opportunity beyond supply availability significant after Nvidia grew revenue 100% year over year. • Infrastructure buildout: Huang said every company will need to participate in the computing shift and build AI infrastructure.