Infineon's CY8CKIT-062S2-AI Dev Kit Targets Low-Power ML and AI at the Edge

Deploy ready-to-run models that tie into a range of on-board sensors — including radar — or roll your own with Imagimob Studio.

Infineon has unveiled an all-in-one development board that aims to deliver a quick-start platform for working with machine learning and artificial intelligence (ML and AI) at the edge: the CY8CKIT-062S2-AI, powered by its PSoC 6 microcontroller unit.

"Infineon's PSoC 6 Artificial Intelligence Evaluation Kit is your perfect partner for easy prototyping and collecting real-life data to build compelling ML products fast," the company claims of its creation. "Being only the size of a cracker, the hardware platform focused on edge AI enables customers to evaluate Infineon's machine learning platform Imagimob Studio, as well as ready to deploy ML models and other software products."

Infineon has unveiled a new all-in-one development kit for those looking to experiment with edge AI — complete with on-board radar. (📷: Infineon)

The CY8CKIT-062S2-AI board, brought to our attention by Linux Gizmos, is based on the company's PSoC 62S2 microcontroller, which includes two different Arm cores on board: an Arm Cortex-M4 running at up to 150MHz and a Cortex-M0+ running at up to 100MHz, combined in a "microcontroller subsystem" that features cryptographic acceleration, a real-time clock (RTC), 8kB of instruction cache, 1MB of static RAM (SRAM), and 2MB of flash memory.

This is connected to a surprisingly comprehensive set of on-board sensors, given the board's compact dimensions: there's a six-axis inertial measurement unit (IMU) and dedicated three-axis magnetometer, a pressure sensor, a pair of digital MEMS microphones, and — most surprising of all — a radar sensor, the Infineon XENSIV 60GHz BGT60TR13C. This, the company explains, includes a finite state machine (FSM) to reduce power draw and processing load on the host microcontroller, while offering human presence detection out to nearly 50 feet with sub-millimeter movement sensing.

Infineon has a range of "Ready Models" which can be deployed on-device as a quick-start, or you can roll your own. (📷: Infineon)

All of this is designed to feed data into an on-device machine learning model — and the company has a range ready to deploy, if you're looking for the quickest start possible. If you'd prefer to roll your own, meanwhile, the kit is designed to link to Imagimob Studio for data-gathering, custom model development, and tuning — and there's on-board Wi-Fi and Bluetooth for wireless communication.

More information on the kit is available on the Infineon website, where interested parties can sign up to be notified when it goes on sale.

ghalfacree

Freelance journalist, technical author, hacker, tinkerer, erstwhile sysadmin. For hire: freelance@halfacree.co.uk.

Latest Articles