Jensen Huang said AI has reached an inflection point, with generated tokens now driving productivity and profit and compute effectively equating to revenue. Demand for compute is accelerating: what was once driven mainly by a single AI lab has become a wave of new AI labs and startups scaling concurrently, open-model ecosystems expanding, and applied AI moving into production. AI infrastructure build-out is running at full speed globally, and NVIDIA’s Vera Rubin cluster is now fully operational

Jensen Huang said AI has reached an inflection point, with generated tokens now driving productivity and profit and compute effectively equating to revenue. Demand for compute is accelerating: what was once driven mainly by a single AI lab has become a wave of new AI labs and startups scaling concurrently, open-model ecosystems expanding, and applied AI moving into production. AI infrastructure build-out is running at full speed globally, and NVIDIA’s Vera Rubin cluster is now fully operational and purpose-built for this moment.