//php echo do_shortcode(‘[responsivevoice_button voice=”US English Male” buttontext=”Listen to Post”]’) ?>
One of the stars of this year’s Infineon OktoberTech 2025 event at the Computer History Museum in Mountain View, Calif., was Ameca, a humanoid robot. In the opening introductions, Infineon Technologies’ CMO Andreas Urschitz interacted and joked with Ameca on stage as if the humanoid robot was just another speaker. The delays in responses made the conversation sound stuttered, but that is because Ameca was taking in the voice inputs it heard, sending the data to the cloud via the venue’s Wi-Fi, and processing an intelligible and intelligent response a few seconds later.
During the morning keynotes, Ameca was positioned at the side of the stage and was “listening” to all the keynotes. Speakers engaged with Ameca briefly, but the piece de resistance was when Infineon’s CMO asked Ameca to summarize the keynotes from the morning (which I thought was quite good, a kind of humanoid robot manifestation of ChatGPT or similar).
The point of the humanoid robot is that it is a part of the physical AI phenomenon that everyone seems to be marketing these days. This essentially translates to robotics and at OktoberTech 2025, you could see some focus on the technologies enabling it. One of the speakers at the event, Deepu Talla, VP and general manager for robotics and edge AI at Nvidia, also heralded this era of physical AI.
“We are in a golden moment for robotics and physical AI,” Talla said, indicating that the technology is now available and that there was a general pursuit of a “robotics brain.”
By Menta 10.30.2025
By GigaDevice 10.29.2025
By BIWIN Storage Technology Co., Ltd. 10.29.2025
“Much of the AI deployed today is extremely brittle, and that’s why we need to get to [developing] a general-purpose robotics brain.”
Adam White, division president for power and sensor systems at Infineon Technologies, said that AI is becoming the backbone of progress across industries, and this is where delivering power efficiently, whether for data centers or at the edge, is a key area of Infineon’s strength. White illustrated the similarities between technologies for humanoid robotics and automotive, especially with central compute, zonal compute, motor control, battery management systems, charging and environmental sensing.
You can watch what Adam White and Deepu Talla said, plus thought from other speakers, at the event at this link here.
What goes on at OktoberTech?
In order to get more insights into OktoberTech and the highlights for 2025, EE Times spoke to Maher Matta, Infineon Americas’ president, and Andreas Urschitz, Infineon Technologies’ CMO. You can watch the video of the conversation below:
In the video, Urschitz talks about the Infineon OktoberTech 2025 events around the world, the importance of the event in Silicon Valley, and some of Infineon’s key achievements and highlights over the last 12 months. And Matta highlights his favorite demos and focus areas from the event this year at the Computer History Museum in Mountain View. In particular, one of Matta’s highlights was the edge AI demo, which featured Infineon’s new Deepcraft AI suite optimized for its PSOC edge microcontrollers for edge AI (see detailed story on this in EDN, “Edge MCUs bolstered by AI design toolchain”).
Edge AI focus
The latter technology, edge AI, has become increasingly significant for many reasons, as Thomas Rosteck, Infineon’s division president of connected secure systems, highlights in this video below:
Here, Rosteck talks about how edge AI is becoming more of a reality as computing performance at the edge also increases. He says that IoT is moving into a new dimension, where it is not just about connecting devices to the cloud, but now providing a useful interface for humans.
He also talks about the Deepcraft AI suite and how it provides a comprehensive development environment to help customers seamlessly integrate AI into their products. The suite will, according to Infineon, provide an edge AI software ecosystem that helps unlock the full potential of edge AI. Developers can access all Infineon’s solutions via its centralized Deepcraft AI Hub and ModusToolbox, allowing them to contribute their models or adopt ready-to-use ones.
The suite also includes Deepcraft voice and audio solutions for the development of high-quality, voice-controlled products to improve speech intelligibility by removing unwanted noise and supporting natural voice interfaces running locally on edge devices.
On PSOC Edge, Infineon said the voice and solutions achieve always-on listening below 1mW with very low latency room conditions, avoiding repeated wake-word prompts and extending battery runtime. Detection rates exceed 98% in close-talking scenarios, with a very low rate of false alarms. Alongside this low-power, always-on listening, PSOC Edge claims to deliver up to 75% faster audio processing at roughly half the energy consumption of competing solutions, which further extends device performance and battery life. The combination of AI acceleration and advanced memory architecture also makes it possible to run multi-megabyte models, such as Voice ID, that would exceed the capabilities of modern DSPs.
Power: Moving to 800-V powered AI data centers
During the week of OktoberTech, the OCP Summit was also taking place in San Jose, Calif. As Majeed Ahmad, editor-in-chief of EDN magazine, wrote in his article, “The exponential growth of AI is rapidly outstripping the capacity of the current 54-V data center power infrastructure, driving a transformation toward high-density, reliable, and safe 800-V powered data centers. Here, at this technology crossroads, the new power delivery architecture requires new power conversion solutions and safety mechanisms to prevent potential hazards and costly server downtimes.”
So, at OCP Summit and Infineon OktoberTech, this tectonic shift in data center power infrastructure was a major highlight. Infineon demonstrated 800-V AI data center power architectures built around silicon, silicon carbide (SiC) and gallium nitride technologies.
At Infineon OktoberTech Silicon Valley, Peter Wawer, division president of green industrial power at Infineon Technologies, spoke with EDN to explain the transition to AI data centers to 800-VDC architectures. He also walked through the demo to show how 800-V power is delivered to AI server racks.
You can watch the video below:
EDN’s article goes on to highlight many other power electronics developments from Infineon, such as next generation of solid-state circuit breakers, a power system built around high-voltage CoolSiC components for high-voltage DC power distribution to IT racks powered by a solid-state transformer, a reference board for hot-swap controllers for 400-V and 800-V power architectures in AI data centers, and trans-inductance voltage regulator modules specifically designed for high-performance AI data centers. Full details are over at EDN at this link.
If you want to watch the videos of all the talks that took place at Infineon’s OktoberTech Silicon Valley 2025 event, check them out here.
See also:
Silicon Valley Events Highlight Rapid Innovation from Power to Edge AI
The transition from 54-V to 800-V power in AI data centers
Edge MCUs bolstered by AI design toolchain