Nvidia’s CEO Jensen Huang, at CES 2026, announced a transformative era for artificial intelligence with the full production launch of the Vera Rubin supercomputing platform. This marks a pivotal moment, as AI transitions into a new computing cycle characterized by applications built directly on AI and a fundamental redefinition of software development. The Vera Rubin platform, succeeding the Blackwell architecture, is poised to deliver unprecedented performance and efficiency, addressing critical challenges in AI's evolution.
Huang highlighted a profound shift in the computing industry, asserting that major transformations occur roughly every 10 to 15 years. This current period is unique, witnessing two simultaneous shifts: the pervasive integration of AI into applications and a complete overhaul of software creation methodologies. Nvidia's commitment to advancing AI was evident in Huang's presentation, which spanned from the scaling of AI capabilities to the development of agentic systems and the integration of physical laws into AI understanding.
The CEO elaborated on the progression of AI beyond conventional chatbots, emphasizing the emergence of 'agentic AI.' These advanced models are designed for autonomous planning, reasoning, and action, moving beyond mere text-based interactions. Huang articulated that information extends beyond language, suggesting that AI can learn from any structured data in the universe. This includes 'physical AI,' which is trained to comprehend and interact with the real world by applying the principles of physics. Nvidia showcased this with its Cosmos world foundation model, which generates realistic simulations for training robots and autonomous systems, including self-driving vehicles.
A significant focus of Huang's keynote was Nvidia's dedication to open AI ecosystems. He noted the rapid advancements of open models, which are now merely six months behind proprietary frontier models and continuously narrowing the gap. A substantial majority of startups are reportedly leveraging open models, and a significant portion of AI usage across developer platforms relies on open-source systems. Nvidia is actively contributing to this trend by releasing its models, along with the data and lifecycle tools essential for their training, evaluation, and deployment. This strategy aims to foster broader innovation and accessibility within the AI community.
Nvidia demonstrated the practical applications of physical AI through various innovations, including the Cosmos world foundation model for robotics and autonomous vehicles. The company also introduced Alpamayo, an open-source AI specifically designed for autonomous driving, enabling vehicles to learn from limited real-world data and navigate complex, unfamiliar scenarios. The integration of advanced AI in robotics was vividly illustrated by the appearance of Star Wars' BDX droids on stage, which are fully autonomous and powered by Nvidia Cosmos. Furthermore, Nvidia's collaboration with Siemens signifies a major advancement in manufacturing, where physical AI utilizes synthetic data from digital factory twins to train next-generation robotics, promising a future where intelligent machines play an even more integral role in industrial processes.
The Vera Rubin platform itself is a technological marvel, offering up to five times the performance of its predecessor, Blackwell, while significantly enhancing efficiency, memory bandwidth, and interconnect speeds. This platform incorporates advanced GPUs, custom CPUs, high-speed networking, and full-stack encryption, all designed to overcome what Huang identified as the next major constraint in AI: context and data movement. The system features NVIDIA's ConnectX-9 Spectrum-X SuperNIC and can be assembled remarkably quickly, in contrast to older systems. Its innovative water-cooling system, operating at around 45°C, not only ensures optimal performance but also dramatically reduces energy consumption. The platform's name honors astronomer Vera Rubin, whose pioneering work led to the discovery of dark matter, symbolizing Nvidia's commitment to groundbreaking scientific and technological exploration.
In summary, Nvidia's latest announcements at CES 2026 underscore a profound evolution in artificial intelligence. With the Vera Rubin supercomputing platform now in full production, Nvidia is not only pushing the boundaries of AI performance but also championing open AI ecosystems and the practical implementation of physical AI in robotics and autonomous systems. The company's strategic vision aims to address the foundational shifts in computing, ensuring AI's continued advancement and widespread adoption.