AEI

ASIA ELECTRONICS INDUSTRYYOUR WINDOW TO SMART MANUFACTURING

NXP, NVIDIA Unveil Physical AI for Humanoid Robotics

At NVIDIA GTC 2026, NXP Semiconductors announced a groundbreaking collaboration with NVIDIA to accelerate the deployment of Physical AI. Specifically, the partnership introduces a new suite of integrated “robot body” solutions designed to streamline data processing and transport in next-generation autonomous machines, including humanoid robots.

The collaboration centers on the integration of the NVIDIA Holoscan Sensor Bridge with NXP’s high-performance System-on-Chips (SoCs). By combining NXP’s expertise in edge processing and functional safety with NVIDIA’s AI infrastructure, the companies have developed a unified architecture that bridges the gap between a robot’s central “brain” and its physical “body.”

NXP announced the first in a series of innovative robotics solutions, developed in collaboration with NVIDIA

Solving the Latency Challenge

Humanoid robots require synchronized motion and dense sensor fusion to operate safely. NXP’s new solutions allow developers to establish a direct, low-latency transport route for real-time data. This architecture significantly reduces software complexity while shrinking the physical footprint and power consumption of robotic systems.

“Physical AI is redefining what machines can do in the real world,” said Charles Dachs, Executive Vice President at NXP. “By combining our expertise in secure networking and real-time control with NVIDIA’s platforms, we are enabling seamless connectivity for the most complex robotic revolutions.”

Product Roadmap and Availability

The first “Holoscan Sensor Bridge-ready” products in NXP’s portfolio include the following:

Machine Vision: A solution based on the i.MX 95 applications processor for high-bandwidth data ingestion.

Precision Motor Control: A system utilizing a kinematic chain of i.MX RT1180 crossover MCUs and the S32J TSN switch for synchronized actuation.

These ready-to-deploy solutions support popular industrial protocols like EtherCAT and Time-Sensitive Networking (TSN). According to NXP, the initial portfolio will be available to the market in the first half of 2026, providing a scalable foundation for full-body robot design.

17 March 2026