Read, Watch or Listen

TAIPEI, Taiwan – Nvidia CEO Jensen Huang used his keynote at the Computex 2026 exhibition in Taipei to announce that the company’s new Vera central processing units are now in full production for data center deployment. The Vera CPUs are designed specifically for large-scale AI data center workloads and are positioned as a primary growth driver for Nvidia as demand increases for AI agents, generative AI models and other autonomous systems. Nvidia said early adoption of the Vera architecture is already underway, with major organizations including Anthropic, OpenAI and SpaceXAI expected to be among the first customers, underscoring strong interest from leading developers of artificial intelligence technologies. TAIPEI, Taiwan – In the same presentation, Huang also unveiled a reference design for a humanoid robot that Nvidia intends as a blueprint for future research and development, particularly in the higher education sector. The company said the reference architecture offers universities and research institutions a standardized foundation on which to build and test advanced robotic systems. Nvidia presented the design as part of its broader effort to provide end-to-end infrastructure for generative AI and autonomous systems, aiming to support the development of physical robots that can integrate AI capabilities and be used in experimental automated labor and other real-world applications.

Prepared by Jonathan Pierce and reviewed by editorial team.