Infineon says it is expanding its collaboration with Nvidia to push system architectures for Physical AI, with humanoid robots in its sights.
The pair are building on a tie-up announced in August 2025, with Infineon leaning on motor control, microcontrollers, power systems and security while NVIDIA supplies AI, robotics and simulation kit.
Infineon says it will join Nvidia’s Halos AI Systems Inspection Lab, aiming to nail down the hardware and software safety foundations so robots behave safely and securely in the real world.
It reckons humanoid robots are nasty bits of engineering that have to perceive their surroundings, make decisions in real time and act safely in workplaces built for humans.
That means a long chain of semiconductor functions, sensing, processing, actuation, connectivity and energy management, and Infineon wants its parts threaded through Nvidia’s robotics stack.
A big chunk of the plan is digital twins for Infineon smart actuators and selected sensors, dropped into NVIDIA Isaac Sim and Nvidia Isaac Lab so developers can tune motion control and perception before touching hardware.
Infineon claims that such simulation-led development surfaces problems early, shortens time-to-market, and reduces integration risk for humanoids aimed at logistics, manufacturing, and service robotics.
Infineon division president Power & Sensor Systems Adam White said: “The rapid expansion of the robotics market will drive growth in the semiconductor market, as intelligent robots rely on microelectronics to sense, think, and act – safely and securely.”
He said Infineon’s power, motor control, microcontrollers and security paired with Nvidia’s digital-twin platforms and the Halos lab should speed validation and push pilots into real operations.
Nvidia vice president of robotics and edge AI Deepu Talla: “The next generation of humanoid robots demands physical AI that can safely navigate complex, real-world environments while meeting the highest standards of functional safety and security,” said.
He said integrating Nvidia Jetson Thor and the simulation tools into Infineon sensing and actuation gives developers a platform to speed the rollout of safe autonomous machines.
On the engineering side, Infineon and Nvidia are talking up a common humanoid robot architecture built for ultra-low latency, compact form factors, and high power density, which reads like a shopping list for heat sinks.
Infineon says it will provide motor-control solutions driven by Nvidia Holoscan Sensor Bridge, interfacing with the Nvidia Jetson Thor developer platform.
The firm is pushing its Infineon AURIX microcontrollers and PSOC devices into the reference design story, with support for post-quantum cryptography for firmware and system protection.
Infineon wants to supply hardware TPM chips and other security components as reference designs to protect AI models and data from the Jetson module all the way out to the cloud.
The collaboration also leans on the Halos safety framework, with Infineon claiming it can help create certifiable systems for Level 4 autonomous vehicles and robotics.
Infineon says it will bring hardware and software safety foundations, tying platforms and operating systems together so safety and cybersecurity are designed in across the stack.
That is supposed to help anyone building Jetson carrier boards bake in secure boot, encrypted communications and safe over-the-air updates, rather than bolting them on after the first scare.
