Nvidia Signals Pivot to Physical AI at CES 2026

Nvidia’s CES 2026 keynote, led by CEO Jensen Huang, marked a departure for the company as the focus shifted to “physical AI,” or the evolution of AI from digital assistants to machines that can think and act in the real world. Here are the key takeaways from the event.

Nvidia provided details on Vera Rubin

Nvidia provided more details about Vera Rubin. The Rubin platform, named after the pioneering astronomer, is the successor to the Blackwell architecture. Rubin is Nvidia’s first “extremely engineered” system where the GPU, CPU (Vera), networking (BlueField-4) and storage are designed as a single fabric. It is designed to reduce the cost of AI tokens to a tenth of the current cost while providing significantly higher throughput per megawatt.

“Rubin comes at exactly the right time as the demand for AI computing for both training and inference is rising,” Nvidia CEO Jensen Huang said in a statement.

He added: “With our annual cadence of delivering the next generation of AI supercomputers – and extreme code design across six new chips – Rubin has taken a giant leap towards the next frontier of AI.”

Commenting on Rubin, William Blair analyst Sebastien Naji said: “Nvidia pointed out that given the significant performance increase in Vera Rubin, it would need roughly a quarter of the number of GPUs. [graphics-processing units] train a model against Blackwell.’

Autonomous driving

Nvidia sees autonomous driving as another growth driver. While this vertical currently only accounts for about 1% of Nvidia’s revenue, it shows a lot of promise. “We envision that one day there will be a billion autonomous cars on the road,” Huang said. He added: “You can either have it as a robot taxi that you organize and rent from someone, or you can own it.”

Nvidia has reaffirmed its goal of deploying Robotaxi Level 4 by 2027. The company is positioning its software stack as a “robotaxis-ready” reference architecture and is encouraging other manufacturers to build on its open source foundation to compete with closed systems like the Tesla FSD.

NVIDIA introduced Alpamayo, a new family of open-source AI models and tools designed to solve the “long end” of autonomous driving. Alpamayo 1 is a 10-billion-parameter Vision-Language-Action (VLA) model that processes video input and generates driving trajectories, but most importantly, it also provides “chain of thought” reasoning that explains why it took a particular action.

For example, if a car hits a broken traffic light, the model might reason that it should follow a traffic officer’s hand signals or standard yield rules rather than just stop indefinitely.

Physical artificial intelligence

Huang sees physical AI as the next big opportunity, saying, “The ChatGPT moment for robotics is here. Breakthroughs in physical AI – models that understand the real world, reason and plan actions – are unlocking entirely new applications.”

This sentiment is shared by Wedbush analyst Dan Ives, who is known for his perma-bullish views on tech stocks. “We believe the robotics and autonomous technology market represents an incremental market opportunity for Nvidia to capitalize on, which is indicative of our view that the company will exceed a market cap of $5 trillion in the near future, and could ultimately be a market cap of $6 trillion,” Ives wrote in his research report.

Nvidia introduced a series of Cosmos models

Nvidia introduced the Cosmos family of foundational models designed to give robots physical intuition. Cosmos can generate realistic videos from individual frames and predict the physical trajectories of objects to help robots “understand” their environment.

The company also unveiled the Isaac GR00T N1.6, which offers improved full-body control for humanoid robots, allowing them to perform complex manipulative tasks while walking. It also unveiled a new robotics module powered by Blackwell that delivers 4x the performance of the previous generation for industrial AI.

Playing

While there was no new hardware, Nvidia pushed the boundaries of AI-driven gaming performance. DLSS 4.5 features a “2nd Gen Super Resolution Transformer” for improved image stability. A new generation of 6x Dynamic Multi-Frame has been announced for the RTX 50 series, targeting 4K 240Hz performance. Also RTX Remix Logic: A new tool for modders that allows classic games to react to in-game events in real-time (like lighting or automatic weather changes when a character opens a door). Nvidia has announced native apps for Linux and Amazon Fire TV, expanding the reach of GeForce NOW.

The main theme of the keynote was Nvidia’s transition from a chip maker to an “Intelligence Backbone”. By open-sourcing massive models like Alpamayo and Cosmos, Nvidia is trying to do for robotics and advance what it did for AI research with CUDA, creating an ecosystem where everyone builds on Nvidia software.

Analysts were generally underwhelmed by the CES 2026 announcements. “NVIDIA’s pivot towards AI at scale and AI systems will help it stay ahead of its rivals,” he said Paolo Pescatorean analyst at PP Foresight, in Las Vegas.

He added: “Alpamayo represents a profound shift for NVIDIA, moving from primarily computing to a platform provider for physical AI ecosystems.

US-China AI War

Meanwhile, the AI ​​battle is on as companies try to gain a competitive edge in the technology. There is also an apparent AI war between the US and China. While the former has blocked companies from exporting high-end AI chips to China, citing possible military use, many, including Nvidia CEO Jensen Huang, believe the policy has failed.

China, which has cracked down on its tech companies, notably Alibaba earlier, is now supporting its tech companies in the midst of an AI war with the US. In February, Chinese President Xi Jinping met with the country’s entrepreneurs, including Alibaba co-founder Jack Ma, at a symposium. Mao’s attendance at the event with Beijing was made all the more important by the fact that the Chinese billionaire was the face of China’s crackdown on tech tycoons the Communist Party believed had become too powerful.

Alibaba is among the Chinese companies that have developed AI chips and even secured a major deal with state-owned telecommunications company China Unicom to supply artificial intelligence (AI) chips for a new data center.