Infineon Unveils PSOC Edge Series for AI

On April 24, 2024, Infineon Technologies AG introduced the new PSOC™ Edge microcontroller (MCU) series, engineered for optimized performance in machine learning (ML) applications. The series features three distinct product lines—E81, E83, and E84—that provide scalability in performance, features, and memory options. These MCUs come equipped with comprehensive design tools and software, enabling developers to swiftly bring new concepts to life and accelerate the introduction of innovative IoT, consumer, and industrial ML-enabled applications to the market.

Steve Tateosian, Senior Vice President of Industrial MCU, IoT, Wireless, and Compute Business at Infineon Technologies, stated, "Next-generation IoT edge devices must achieve higher performance while maintaining low power consumption. Infineon's PSOC™ Edge E8 series capitalizes on machine learning capabilities for real-time AI, achieving an optimal balance of performance and energy efficiency. These semiconductors provide embedded security for connected home devices, wearables, and industrial applications, reflecting our dedication to enhancing future IoT systems."

The new PSOC™ Edge E81, E83, and E84 MCUs leverage high-performance Arm® Cortex®-M55 cores, supporting the Arm® Helium™ DSP instruction set and Arm® Ethos™-U55 neural network processor. They also incorporate Cortex®-M33 cores with Infineon's ultra-low power NNLite neural network hardware accelerator. All three lines offer a diverse range of peripherals, on-chip memory, robust hardware security features, and various connectivity options, such as USB HS/FS with built-in PHY, CAN bus, and Ethernet. Additionally, these MCUs facilitate connectivity with Wi-Fi® 6, Bluetooth® Low Energy (BT/BLE), and the Matter protocol.

PSOC™ Edge E81 integrates Arm® Helium™ DSP technology and Infineon's NNLite neural network accelerator. PSOC™ Edge E83 and E84 incorporate Arm® Ethos™-U55 micro NPU processors, providing a 480-fold improvement in ML performance compared to existing Cortex®-M systems. These MCUs also support Infineon's NNLite neural network accelerator, ideal for ML applications in low-power computing.

The target applications for the PSOC™ Edge E8x series encompass human-machine interfaces (HMI) in home appliances and industrial equipment, smart home and security systems, robotics, and wearables. All three series offer voice and audio sensing capabilities for activation and control, with the E83 and E84 providing enhanced features for advanced HMI, such as ML wake-up, visual positioning detection, and facial and object recognition. The PSOC™ Edge E84 series further expands its functionality with low-power graphical displays supporting up to 1028x768 resolution.

The PSOC™ Edge E8 MCUs provide a comprehensive system and software-compatible product line for designers. Hardware design support includes evaluation boards with Arduino expansion interfaces, sensor kits, BLE connections for configuration, and Wi-Fi for smartphone and cloud connectivity.

The PSOC™ Edge series is supported by Infineon's ModusToolbox™ software development platform, which offers a robust suite of development tools, libraries, and embedded runtime environments. This versatile platform supports a wide range of applications across consumer IoT, industrial, smart home, and wearable devices.

Imagimob Studio is an edge AI development platform integrated with ModusToolbox™, offering end-to-end ML development capabilities from data input to model deployment. Starter projects and ready-made models from Imagimob provide an accessible entry point for users. In conjunction with PSOC™ Edge, Imagimob facilitates the rapid construction and deployment of the latest ML models at the edge.