Nvidia’s Huang Visits Taiwan as Rubin Shipments Near

Nvidia is moving its data-center roadmap toward the Vera Rubin architecture, designed to support increasingly demanding generative and agentic AI workloads. Taiwan Semiconductor Manufacturing Co. is central to the effort through advanced wafer production and the packaging and testing capacity needed for volume shipments. The delivery schedule matters because delays could ripple through Nvidia’s supply chain and cloud providers’ plans for their next wave of AI infrastructure.

Nvidia Chief Executive Jensen Huang was recently spotted in Taipei in an unannounced visit, prompting speculation that he was checking progress on Vera Rubin wafer starts, packaging, testing and deliveries at TSMC. Products based on the new architecture are expected to begin shipping in volume to major cloud operators from the third quarter. Nvidia and TSMC have not disclosed Huang’s itinerary, the specific production milestones discussed or any transaction value tied to the visit.