Hardware Podcasts

Chip Combines Analog and Digital Neurons for Sensor Data

EE Times Current · 8 Nov 2024 · 48m

What this episode covers

Sumit Kumar, CEO, and engineers Peter Bogdan, Vasili Toma, and Amir Zeo of Innatera, a spinout from the University of Delft, discuss their neuromorphic processor chip designed for low-power, always-on sensor applications. The chip integrates both analog and digital neurons with a RISC-V CPU to handle pattern recognition in devices like smartwatches and security cameras. The team emphasizes a modular architecture and a PyTorch-based software stack to make the technology accessible to standard ML engineers without requiring a neuromorphics PhD. Professor Ralph Etienne-Cummings of Johns Hopkins University provides commentary on the chip's potential in biomedical applications and the importance of standardization.

Most relevant to

early-stage companies

Topics

Mentioned

Companies
Innatera, Nvidia
Products
Jetson GPU
Chips
RISC-V
Tools
PyTorch
Also
University of Delft, Spinnaker, Johns Hopkins University, Czech Technical University

Related episodes from other shows

Other shows on the same ground, closest first.

  • Embedded.fm · 1h 24m

    497: Everyone Likes Tiny

    Kwabena Agyeman, President and Co-Founder of OpenMV, discusses the launch of two new programmable camera systems, the OpenMV Cam N6 and the OpenMV AE3, via a Kickstarter campaign. He explains how integrating Neural Processing Units (NPUs) and Arm Helium SIMD instructions enables high-performance computer vision and edge AI on microcontrollers, achieving significant speedups while maintaining ultra-low power consumption. Agyeman details the software ecosystem, including MicroPython, TensorFlow Lite for Microcontrollers, and partnerships with Edge Impulse and Roboflow, which simplify model training and deployment for embedded developers.

    Embedded firmwareMicrocontrollers & dev platforms
    8 chapters

  • The Circuit · 31m

    Episode 14: The Wide World of Sensors

    Ben Bajarin and Jay Goldberg discuss the resurgence of the sensor market, driven by the transition from analog to digital and the falling costs of semiconductor manufacturing. They highlight how companies like Texas Instruments, NXP, and STMicroelectronics are leveraging legacy analog capacity and RISC-V architectures to create high-performance, low-cost sensors. The hosts argue that advances in AI and software analytics have finally made the value of this data clear, turning sensors from commodities into strategic assets for industrial IoT and automotive applications.

    Sensors & measurement in productsConnected products & IoT
    8 chapters

  • The Next Byte Podcast · 23m

    110. The Future of Neurology: Neural Interfaces Countering Neural Diseases

    Hosts Daniel and Forbode discuss NeuralTree, a neurochip developed by researchers at EPFL (École Polytechnique Fédérale de Lausanne) designed to detect and counter neural disorders like epilepsy and Parkinson's disease. The chip features 256 input channels, an area of 3.5 square millimeters, and an energy-aware machine learning algorithm that allows it to operate efficiently on the edge without cloud dependency. The hosts highlight the device's ability to distinguish between healthy and irregular neural impulses, its potential for self-evolving algorithms to adapt to changing brain signals, and its broader implications for neurological research and treatment.

    Embedded firmwareMicrocontrollers & dev platforms
    7 chapters

  • Embedded.fm

    27: You Are Blowing My Mind

    Ivo Stivoric, co-founder of Body Media (now part of Jawbone), discusses the shift from simple pedometers to multisensor wearable systems that fuse data from motion, skin temperature, and heat flux to provide accurate health insights and behavior modification. Sam Guilaume and David Rothenberg of Movea explain their indoor navigation algorithm, which uses smartphone inertial sensors for dead reckoning at ultra-low power, and their strategy to harden parts of the sensor fusion logic into ASICs for efficiency. Stephen Walsh of ISKN presents iSketchnote, a MEMS-based pen and paper digitizer that captures handwriting and sketches in real-time, highlighting its successful Kickstarter campaign and transition from prototype to manufacturing.

    Sensors & measurement in productsEmbedded firmware
    7 chapters

  • Moore's Lobby · 53m

    AI or 5G? EdgeQ Asks, "Why Not Both?" While Building a Wireless Base Station-on-a-Chip

    Adil Kidwai, VP and Head of Product Management at EdgeQ, explains how the company built a software-defined base station-on-a-chip that combines 5G and AI capabilities using RISC-V processors. He details the mathematical similarities between information theory and AI, allowing EdgeQ to offer customizable features like low latency or high bandwidth to reduce power consumption and cost compared to traditional multi-chip solutions. Kidwai also reflects on his career transition from Intel's wireless division to a startup environment, highlighting the challenges of scaling during the pandemic and the shift in cellular networks from human-centric to machine-centric communication.

    Microcontrollers & dev platformsWireless & RF
    7 chapters

The summary, chapters and topics are generated from the episode’s audio by this directory, not written by EE Times Current. The episode itself belongs to its publisher.