Infineon Technologies details next-generation PSoC Edge AI microcontroller

Edge devices are evolving rapidly and need to handle increasingly complex tasks, thus requiring more computing power, for which machine learning (ML) offers a compelling solution.

In this article, we will describe how Infineon’s PSoC Edge, a family of high-performance microcontrollers (MCUs) enhanced with machine learning, low power, security, and advanced HMI, can help designers create outstanding user experiences for next-generation devices.

Infineon's new PSoC Edge E81, E83 and E84 MCUs are based on the high-performance Arm Cortex-M55 core, supporting the Arm Helium DSP instruction set and paired with the Arm Ethos-U55 neural network processor, as well as the Cortex-M33 core paired with Infineon's ultra-low power NNLite (a proprietary hardware accelerator for accelerating neural networks). In addition, all three series support a rich set of peripherals, on-chip memory, powerful hardware security features and a variety of connection peripheral options, including USB HS/FS with built-in PHY, CAN bus, Ethernet, support for WiFi 6, BT/BLE connection and Matter protocol, etc. PSoC Edge E81 uses Arm Helium DSP technology and Infineon's NNLite neural network (NN) accelerator. PSoC Edge E83 and E84 have built-in Arm Ethos-U55 micro NPU processors, which have a 480-fold improvement in machine learning performance compared to existing Cortex-M systems, and they support Infineon's NNlite neural network accelerator for machine learning applications in the field of low-power computing.

The target applications of the PSoC Edge E8x series include human-machine interfaces (HMIs) in home appliances and industrial equipment, smart home and security systems, robots, and wearable devices. All three series support activation and control through voice/audio sensing, and the E83 and E84 MCUs provide enhanced features for the implementation of advanced HMIs, including machine learning wake-up, visual position detection, and face/object recognition. The PSoC Edge E84 series also adds low-power graphics display (up to 1028x768) to the rich feature set.

What is PSoC?

PSoC (Programmable System on Chip) is a versatile and well-known microcontroller brand that has been in the market for more than 20 years. PSoC microcontrollers are known for their low power consumption and are widely used in consumer and Internet of Things (IoT) applications. They offer a strong feature roadmap including artificial intelligence (AI), security, integrated connectivity, and programmability.

Omar Cruz, a technology expert at Infineon Technologies, said: "PSoC is a very successful microcontroller brand that has been around for more than 20 years. We are known for our low-power microcontrollers, which are widely used in the consumer and IoT fields."

How is PSoC different from other microcontrollers?

From the beginning, PSoC has focused on reducing BOM through integration. Unlike traditional microcontrollers that only include a processor, memory, and interfaces, PSoC microcontrollers offer much more. Key differences include:

Programmability and Customization: PSoC microcontrollers have programmable analog and digital blocks, allowing customization.

Analog Modules: These modules facilitate the interface between analog sensors and microcontrollers.

CAPSENSE technology: Provides capacitive touch functionality.

Ultra-low power consumption: Designed specifically for battery-powered applications, PSoC microcontrollers excel in low power consumption.

Integrated connectivity: Support includes Bluetooth Low Energy (BLE) and Wi-Fi, which is critical for modern applications.

Human Machine Interface (HMI): PSoC microcontrollers are known for their best-in-class HMI capabilities.

Cruz elaborated: “We have been reducing the BOM through integration. We not only have processors, memory and interfaces, but also programmability and customization, including analog modules and capacitive touch technology.”

The Origin of PSoC

PSoC started out as a peripheral microcontroller designed to add specific features, such as capacitive touch sensing. Over time, it evolved to include more computing power and integrated connectivity. The introduction of PSoC 4 and PSoC 6 brought additional computing power and connectivity options to meet the growing customer demand for more performance and edge computing.

Cruz reflects on the origins of PSoC: “From the beginning, we focused on adding peripheral functionality, such as capacitive touch sensing. Over time, we introduced PSoC 4 and PSoC 6, which brought more computing power and connectivity options.”

Applications of PSoC

PSoC microcontrollers are widely used in various fields:

Smart Home and Consumer IoT: PSoC microcontrollers are a common choice in home appliances with touch interfaces and are widely used in a variety of smart home devices.

Medical Equipment: Ultra-low power capabilities make PSoC ideal for medical equipment that requires reliability and efficiency.

Automotive Infotainment Systems: In automotive systems with touch interfaces, PSoC microcontrollers provide reliable and responsive control.

“We are very well known in the smart home and consumer IoT space, with PSoC microcontrollers in many home appliances,” Cruz said. “We also have a broad presence in medical devices and automotive infotainment systems.”

Challenges Solved by PSoC Edge

PSoC Edge is designed to address several key challenges in modern edge computing and machine learning applications:

Computing performance: Enhanced performance is required to handle machine learning operations at the edge.

Low power consumption: This is critical for battery-powered applications, ensuring that the device can operate efficiently without frequent battery changes.

Security: By processing data at the edge, PSoC Edge reduces the need for data transfer between the device and the cloud, thereby enhancing security.

Integration: Integrate various interfaces such as voice, vision, and sensors into one microcontroller to simplify development and functionality.

More memory: As applications become more complex, the need for more memory is also met by PSoC Edge.

“We addressed several key challenges in PSoC Edge, including computing performance, low power, security, integration, and memory requirements,” Cruz elaborated.

PSoC Edge and Software

Hardware advancements require strong software support. PSoC Edge integrates seamlessly with software tools such as ModusToolbox, providing drivers, middleware, and code examples to interact with the hardware accelerator provided by Ethos-U55. In addition, Infineon's acquisition of Imagimob Solutions further enhances its machine learning capabilities, providing end-to-end ML solutions from data import to deployment. This combination makes it possible to optimize machine learning models and quickly go to market, providing support for production-ready solutions.

Cruz explained: “We not only provide hardware, but also provide strong software support through ModusToolbox, and acquired Imagimob to enhance our machine learning capabilities and provide an end-to-end solution.”

PSoC 6 in detail

PSoC 6 is a product launched in 2017 with higher computing power and more connectivity options. PSoC 6 is designed for applications that require high performance and low power consumption and is suitable for a variety of complex edge computing and IoT devices. Its main features include:

Dual-core architecture: PSoC 6 features a dual-core architecture with one Arm Cortex-M4 and one Arm Cortex-M0+ core, providing flexible performance and power management.

Security features: PSoC 6 integrates hardware cryptographic accelerators, Trusted Platform Module (TPM), and secure boot capabilities to ensure data security and integrity.

High-performance connectivity: Supports BLE 5.0 and Wi-Fi, providing fast and reliable wireless connections.

Rich I/O interfaces: including a variety of analog and digital interfaces, supporting a wide range of sensor and peripheral connections.

Low power design: Optimized power management and ultra-low power modes for battery-powered applications.

The latest PSoC Edge

PSoC Edge is designed for the next generation of responsive computing and control applications, with hardware-assisted machine learning (ML) acceleration. This new MCU series makes terminal products smarter and more intuitive by lowering the threshold for human-computer interaction and adding contextual awareness to terminal applications, thereby providing a higher level of terminal user experience. At the same time, it provides strong privacy and security protection through embedded edge protection technology.

PSoC Edge devices feature the high-performance Arm Cortex-M55, including Helium DSP support paired with Arm Ethos-U55 and Cortex-M33, and Infineon's ultra-low-power neural network accelerator. The lead devices use ultra-low-power 22nm embedded RRAM technology. These devices support "always-on" sensing and response, and are ideal for advanced IoT devices such as smart homes, security, wearables, robots, etc. The PSoC Edge family of devices has ample on-chip memory to support the most advanced applications. PSoC Edge consists of dozens of devices that provide scalability for increasingly complex next-generation intelligent systems while enabling software reuse and portability.

The PSoC Edge family adds intelligence by adding support for advanced graphics, audio, and vision applications to the broad range of human interface modes (touch, graphics, voice) available in previous generation PSoC devices. The new family enables easy migration of applications within the family and provides an upgrade path for previous generation designs.

PSoC Edge has a variety of potential application areas: