Infineon Technologies has developed a family of microcontrollers for energy efficient machine learning and AI inference at the edge with low power resistive RAM (RRAM).
The PSoC Edge devices are based on high performance ARM Cortex-M55, including Helium DSP support paired with ARM Ethos-U55 and Cortex-M33 paired with Infineon’s ultra-low power NNLite and RRAM memory from TSMC. NNLite is a proprietary hardware accelerator intended to accelerate the neural networks used in Machine learning and AI applications.
Support for “always-on” sensing and response makes the devices suitable for advanced IoT and industrial segments such as smart home, security, wearables, robotics, and many more. The family of devices has scalable on-chip memory, including non-volatile RRAM, as well as high speed, secured external memory support.
“Infineon’s new PSoC Edge portfolio is built on years of experience delivering solutions to intelligently control, interact, and connect our world with highly usable and powerful MCU-based systems,” said Steve Tateosian, Senior Vice President of Microcontrollers of Infineon.
“They offer an ideal mix of scalable power and performance to support emerging AI/ML requirements, extensive HMI capabilities and security features required for next-generation applications. All supported with a deep ecosystem of software and tools, fit for task.”
The performance of the PSoC Edge family extends support for advanced graphics, voice, audio, and vision-based applications to the existing PSoC portfolio for easy migration of applications within the family, as well as an upgrade path from existing designs.
Infineon’s ModusToolbox software platform provides a collection of development tools, libraries, and embedded runtime assets for a flexible and comprehensive development experience. ModusToolbox supports a wide range of use cases including consumer IoT, industrial, smart home, wearable, and many other applications.
Imagimob Studio is an Edge AI development platform, integrated into ModusToolbox, and delivers end to end ML development, from data in to model deployed. Starter projects and Imagimob’s Ready Models make it simple to get started. Together with PSoC Edge, Imagimob makes it possible to deploy state of the art machine learning models for the edge.
The PSoC Edge family with RRAM for edge AI is available for early access customers now.