Physical AI was a prominent highlight of this year’s OktoberTech Silicon Valley in Mountain View, California. As AI data centers enter a maturing stage, physical AI is taking the spotlight at the industry events with robotics as a major use case. Infineon managers called this transition a journey from automation to autonomy.
Physical AI moves artificial intelligence from the digital world into machines, factories, and logistics networks. These machines sense, think, and act autonomously in real time, creating unprecedented opportunities in areas ranging from everyday convenience to healthcare, mobility, and industrial automation.
Figure 1 Robots can use physical AI to perceive their environment, communicate securely, process data locally through edge AI, and act safely and reliably while using energy as efficiently as possible. Source: Infineon
At OktoberTech Silicon Valley, Infineon showcased live robots and their key functional blocks in sensing, computing, connectivity, motor control, and power management. Developer booths displayed designs using physical AI to enable next-generation applications, from industrial automation and home appliances to smart home security.
When asked about physical AI’s presence beyond robotics, Stephan Zizala, division president of edge systems at Infineon, mentioned motor control. “Several machine learning algorithms are now powering motor control designs.” He also called robotics the tip of the iceberg for physical AI.
OktoberTech Silicon Valley 2026 displayed system-level solutions for physical AI. Source: Infineon
While talking about the journey from generative AI to agentic AI to physical AI, Zizala called physical AI a new era in artificial intelligence. Physical AI is already at work, but the bigger question is whether it’s ready from a security standpoint, he added. “While physical AI translates perception into precise movement, functional safety and cybersecurity embedded at the hardware level provide an anchor of trust for even the most advanced robots.”
Earlier this year, Infineon unveiled the OPTIGA Trusted Platform Module (TPM), SLB 9672, which secures Nvidia’s Jetson Thor platform. This hardware-based solution securely stores cryptographic keys and verifies system integrity at the chip level, establishing a certified, quantum-resilient root of trust for physical AI systems.
Zizala also discussed the design value chain in which developers take sensor data, train models, and deploy them on microcontrollers. In his keynote, Infineon’s chief marketing officer, Andreas Urschitz, called these MCUs foundational MCUs. In other words, these MCUs are reimagined for edge AI and physical AI applications.
Regarding these AI-era MCUs, developers at OktoberTech Silicon Valley emphasized the need for more memory. It all comes down to more memory, so more AI models could be used, one developer said.
It seems that the data center’s “memory wall” story will come to physical AI designs sooner or later.
Related Content
- Building Physical AI Systems That Can Work in the Real World
- Physical AI Needs a Neuromorphic Path from Sensor to Silicon
- Missing sense: Why physical AI can see everything and feel nothing
- Physical AI Isn’t Just Bigger AI; It’s a Systems Architecture Challenge
- From edge AI to physical AI in smart factories: A shift in how machines perceive and act
The post Physical AI in action at OktoberTech Silicon Valley 2026 appeared first on EDN.