Nvidia unveils 'AI factories' roadmap at GTC Taipei, doubling down on edge and robotics

Nvidia is using GTC Taipei to widen the frame around its AI business. In company materials published ahead of the June 1-5 event, Nvidia said the Taipei program will cover “AI factories,” scaling infrastructure, agentic AI, physical AI and robotics, alongside the accelerated computing products that made it central to the current AI buildout. That wording matters because it shifts attention from the narrow question of who buys the next training cluster to a broader one: who builds, deploys and operates AI systems in factories, warehouses, hospitals and other physical settings. Nvidia’s own event pages pair the data-center language with robotics, edge computing and industrial sessions, and Jensen Huang’s keynote billing says he will unveil the latest in AI, robotics and accelerated computing on June 1 in Taipei. (blogs.nvidia.com) ### Why is Nvidia talking about “AI factories” instead of just chips? Nvidia’s published GTC Taipei agenda uses “AI factories” as an organizing term for the event, alongside scaling infrastructure and agentic AI. The company has used the same phrase in other recent materials to describe large-scale systems that generate tokens, models or AI services as an industrial output rather than as isolated compute projects. (nvidia.com) Google Cloud and Nvidia said in an April company blog post that customers could build AI factories with Nvidia’s Rubin-based systems and related software, showing how Nvidia is tying chips, networking, models and cloud infrastructure into a single stack. That makes the Taipei message less about a one-off slogan than about extending a framework Nvidia has already been using across partners and product launches. (blogs.nvidia.com) ### Where do robotics and physical AI fit into the Taipei event? Nvidia’s event pages say GTC Taipei will cover physical AI and robotics, and community guides posted on Nvidia forums list dedicated “Physical AI Days” with sessions on robot factories, industrial simulation, edge AI and humanoid robots. Those listings show robotics is not a side track at the conference but part of the main agenda. (blogs.nvidia.com) Nvidia’s broader 2026 GTC messaging has also linked physical AI to digital twins, simulation and industrial deployment. In a March company post, Nvidia said robots, vehicles and factories were moving from isolated use cases to enterprise workloads, supported by models including Cosmos 3 and Isaac GR00T N1.7. ### What does the edge push look like in practice? (nvidia.com) GOWIN said on May 21 that it had been invited to Nvidia’s APAC Robotics and Edge AI Partner Day during GTC Taipei. The company described the event as part of Nvidia’s regional ecosystem activity around robotics and edge AI, offering one concrete sign that Nvidia is convening partners beyond hyperscale cloud operators. The GOWIN statement is a company release, but it aligns with Nvidia’s own public emphasis on robotics and edge computing in Taipei. (blogs.nvidia.com) Nvidia’s forum guides also list a June 4 session titled “How to Run Open Models and Agentic AI Efficiently at the Edge,” underscoring that edge deployment is part of the planned programming rather than an afterthought. ### Who shows up if this expands beyond hyperscalers? Taipei-based manufacturing and electronics groups are already prominent in Nvidia’s Taiwan messaging. (blogs.nvidia.com) Nvidia has separately highlighted work with Delta Electronics, Foxconn, TSMC and Wistron on digital twins and industrial AI in Taiwan, tying local manufacturing partners to its physical AI narrative. (forums.developer.nvidia.com) The keynote page also lists participants tied to Taiwan’s infrastructure and manufacturing ecosystem, including Pegatron, Foxconn, National Taiwan University, Visionbay.ai and GMI Cloud. That lineup suggests Nvidia is using Taipei to speak to system builders, manufacturers and regional partners as much as to cloud buyers. ### What should readers watch next? June 1 is the next hard date. (blogs.nvidia.com) Nvidia says Huang will deliver his GTC Taipei keynote at 11 a.m. Taiwan time at Taipei Music Center, and the broader event runs through June 5 with sessions on robotics, edge AI, industrial simulation and AI infrastructure. Nvidia’s conference pages say sessions will be available through its event platform after the show. (nvidia.com 1) (nvidia.com 2)