Arm Powers Physical AI Revolution at CES 2026 Against Qualcomm Robotics Push

At CES 2026, the technology landscape has shifted toward machines that can sense, plan, and act in the real world. Arm is leading this transition by positioning its architecture as the essential foundation for physical AI. This phase of development moves intelligence out of the cloud and directly into the edge devices that surround us. A primary driver of this shift is the need for extreme energy efficiency and local processing power. Arm’s newest automotive-enhanced platform, the Neoverse V3AE, provides the server-class performance necessary for safety-critical tasks. It allows autonomous systems to make split-second decisions without relying on a distant data center.

Qualcomm has entered the fray with its own specialized hardware, the Dragonwing IQ10 robotics processor. This chip is designed specifically for industrial robots and humanoid systems, offering a highly integrated solution for manufacturers. It leverages Qualcomm’s expertise in mobile efficiency to power machines that must operate for long periods on battery power.

While both platforms are built on Arm technology, they represent different paths for the industry. Qualcomm offers a more vertical, all-in-one approach for robotics teams wanting a fast route to market. Arm provides the horizontal scale that supports partners like NVIDIA, whose Jetson Thor hardware also powers advanced robotics.

Major players like Rivian and Mercedes-Benz are already standardizing on these architectures to build their next generation of vehicles. These AI-defined machines utilize Arm to manage everything from infotainment to fully autonomous driving stacks. This creates a unified ecosystem where software can be shared across different types of physical machinery. The emergence of humanoid robots like the Agibot A2 demonstrates how far this technology has come. These machines require a common compute standard to ensure that complex actions like walking or dancing can be programmed effectively. By providing a stable platform, Arm and its partners are making the deployment of physical AI a commercial reality.

CES 2026 has become a defining moment for the semiconductor industry as Artificial Intelligence moves from the cloud into the physical world. Arm is leading this transition by providing the compute foundation for "Physical AI," a category of systems that perceive, plan, and act autonomously. Major hardware launches from industry leaders demonstrate that Arm based platforms are now the de facto choice for powering intelligent robots, autonomous vehicles, and sophisticated edge devices. NVIDIA CEO Jensen Huang described this shift as the ChatGPT moment for physical AI, unveiling the Jetson Thor computer built on Arm Neoverse cores. This hardware is designed to run open robot foundation models that allow machines to adapt to dynamic environments in real time. By leveraging Arm's architecture, NVIDIA is able to provide the petaflop-class performance required for robots to reason like humans while maintaining the necessary efficiency for mobile operation. Qualcomm is challenging this space with its own new entry, the Dragonwing IQ10 robotics processor. While NVIDIA focuses on high-end simulation and reasoning, Qualcomm is targeting the industrial and humanoid sectors with a focus on extreme energy efficiency at the edge. The Dragonwing IQ10 runs on the Arm compute platform, proving that even the most competitive rivals in the chip space are standardizing on Arm's instruction set for their robotics portfolios.

This comparison reveals a strategic split in how physical AI will be deployed. NVIDIA’s approach prioritizes raw processing power for complex planning, while Qualcomm emphasizes low-latency, long-battery-life operation for mobile autonomous systems. Despite these different goals, both companies rely on the Arm ecosystem to ensure their hardware can integrate with the existing software frameworks used by developers globally.

The shift to Arm-based physical AI is already appearing in high-profile consumer products. The Lucid Gravity SUV debuted at the show featuring Level 4 autonomy powered by the NVIDIA DRIVE Thor platform and Arm Neoverse V3AE technology. This demonstrates that the same architecture managing data center inference is now robust enough to handle safety-critical driving decisions on the road. Beyond individual chips, the ecosystem scale of Arm provides a common ground for software portability. This allows a developer to design a physical AI model on a deskside system and deploy it directly to a factory robot or a consumer vehicle. As AI becomes embedded in the world around us, the ability to scale intelligence efficiently from the cloud to the edge is becoming the ultimate competitive advantage.