Friday, April 19, 2024

The Ultra-Low Power Neuromorphic Microcontroller

- Advertisement -

The technology mimics the brain’s processing of sensory data using spiking neural networks, revolutionizing sensor data processing and enabling applications in wearables, smart homes, and IoT devices.


Innatera launched its neuromorphic microcontroller, the Spiking Neural Processor T1. The Spiking Neural Processor is a neuromorphic processing technology that mimics the brain’s mechanisms for processing sensory data. It uses a mixed-signal computing architecture to implement spiking neural networks (SNN), a breed of event-driven neural networks suited for pattern recognition and signal processing in noisy, time-series data.

The T1 system-on-chip is set to transform sensor data processing at the edge, opening possibilities for applications in wearables, smart homes, and IoT devices. With its event-driven processing capabilities, the T1 enables gains for always-on sensing use cases. The T1 integrates mixed-signal neuromorphic computing technology with a RISC-V processor and supports accelerating traditional CNN models.

- Advertisement -

A Talamo software development kit complements the Spiking Neural Processor. Integrated with the PyTorch framework, Talamo provides a platform for developing and deploying spiking neural network applications. The SDK is a tool for developers looking to adopt neuromorphic processing for their applications.

“Neuromorphic computing is here and will redefine intelligence at the sensor edge. We’re excited to unveil the Spiking Neural Processor and announce the availability of the T1 to customers for pre-production trials.”, said Sumeet Kumar, CEO at Innatera. T1 evaluation kits are available as part of Innatera’s early access program ahead of mass production later this year.

The technology enables rapid pattern recognition with sub-1mW power consumption, utilizing spiking neural networks. Through event-driven spiking neural networks, it achieves unmatched power performance in signal processing and pattern recognition tasks.

The device boasts versatile processing capabilities, efficiently implementing spiking neural networks (SNNs), deep neural networks (DNNs), and conventional processing within a single unit.

The device features a variety of interfaces, including QSPI, I2C, UART, JTAG, GPIO, and a front-end ADC. It is also compact, measuring just 2.16mm x 3mm and housed in a 35-pin WLCSP package.

For more information, click here.

Nidhi Agarwal
Nidhi Agarwal
Nidhi Agarwal is a journalist at EFY. She is an Electronics and Communication Engineer with over five years of academic experience. Her expertise lies in working with development boards and IoT cloud. She enjoys writing as it enables her to share her knowledge and insights related to electronics, with like-minded techies.


Unique DIY Projects

Electronics News

Truly Innovative Tech

MOst Popular Videos

Electronics Components