During the GTC Taipei conference, NVIDIA Chief Executive Officer Jensen Huang announced several new NVIDIA AI platforms and partnerships to expand global computing infrastructure. These developments aim to support the deployment of agentic and physical artificial intelligence across multiple industries. Consequently, the company is expanding its hardware and software offerings from cloud systems to personal computers.
New Hardware and Architecture
Specifically, the company announced the full production phase of its Vera Rubin architecture designed for AI factories. In addition, NVIDIA introduced the Vera central processing unit alongside the Vera BlueField-4 STX platform for secure storage processing. Meanwhile, the new DGX Station brings high-performance computing capabilities directly to corporate office environments.
Autonomous Vehicles and Robotics
For the transportation sector, the DRIVE Hyperion system has transitioned into a global platform for autonomous robotaxi services. Furthermore, the company released Alpamayo 2, which serves as an advanced reasoning model for self-driving vehicles. Notably, these technologies integrate with the newly launched Cosmos 3 foundation model designed for physical artificial intelligence applications.
Expanding the NVIDIA AI platforms
To support developers, the company released a collection of open-source tools and the Isaac GR00T humanoid robot reference design. These resources allow researchers to build physical AI systems and deploy them in industrial settings. As a result, these NVIDIA AI platforms provide a structured roadmap for organizations building modern digital factories.
Industry Partnerships and Cloud Infrastructure
In terms of manufacturing, NVIDIA is collaborating with TSMC to integrate artificial intelligence into semiconductor fabrication facilities. Moreover, a partnership with Foxconn and Taiwanese medical centers will introduce these technologies to healthcare systems. Consequently, the global cloud network is expanding to meet the rising demand for specialized computing resources. Meanwhile, Microsoft is working with the company to optimize Windows personal computers for local AI tasks.
Ultimately, these software and hardware releases represent a shift toward agent-based computing models. Enterprise software providers are already utilizing NVIDIA AI platforms to build custom digital agents for business operations. Therefore, these developments establish a foundation for the next phase of global industrial automation. These systems will continue to roll out to global markets throughout the year.
