Edge MCUs bolstered by AI design toolchain

Edge AI designs, starting to see a trickle-down effect from AI data centers, increasingly rely on toolchains to keep up with the breakneck speed of AI. So, to bolster the edge AI ecosystem, Infineon Technologies has expanded its edge AI toolchain with the DEEPCRAFT Suite, a set of software, tools, and solutions that help engineers seamlessly integrate AI into their designs.

DEEPCRAFT AI Suite includes an AI Hub with ready models and audio tuning tools. That simplifies the implementation of AI/ML capabilities in edge devices and allows design engineers to either develop their models from scratch or integrate off-the-shelf models.

Figure 1 DEEPCRAFT AI allows developers to bring their own model and convert it for the edge. Source: Infineon

“With the introduction of our DEEPCRAFT AI Suite, we are further expanding Infineon’s Edge AI software ecosystem for unlocking the full potential of edge AI,” said Steve Tateosian, senior VP and GM for IoT, consumer, and industrial MCUs at Infineon.

Take AI Hub, for instance, which Infineon calls a one-stop shop for its Edge AI software offerings. It offers access to more than 50 content resources, including open-source models, Infineon software, tools, and solutions, as well as case studies from industrial, consumer, and automotive applications.

Then there is DEEPCRAFT Studio, which provides support for audio, computer vision, radar, and other time-series data. It facilitates an end-to-end platform for developing robust AI and machine learning models for use at the edge.

Figure 2 DEEPCRAFT Studio includes training and deploying high-performance computer vision models for object detection using advanced YOLO models. Source: Infineon

Additionally, DEEPCRAFT Model Converter in the suite allows developers to optimize both proprietary and open-source models to run on Infineon hardware. It supports popular AI frameworks, including PyTorch, TFLite, and Keras.

Figure 3 This software tool converts, optimizes, and validates AI models to run on the edge. Source: Infineon

Voice and audio solutions in the DEEPCRAFT suite support the development of high-quality, voice-controlled products. These solutions feature always-on listening below 1 mW with very low-latency room conditions, avoiding repeated wake-word prompts and extending battery runtime. Moreover, detection rates exceed 98% in close-talking scenarios with a very low rate of false alarms.

More specifically, DEEPCRAFT Audio Enhancement improves speech intelligibility by removing unwanted noise. Furthermore, DEEPCRAFT Voice Assistant supports natural voice interfaces running locally on edge devices.

The DEEPCRAFT AI Suite is optimized for Infineon’s PSOC microcontrollers—built around Arm Cortex-M processor cores—to facilitate high-performance, low-power, and secure hardware with machine learning (ML) acceleration in edge applications.

PSOC microcontrollers also provide advanced security features, including Infineon Edge Protect Category 4 (EPC4) with PSA Certified L2 and L4 iSE, PCI pre-certification, and a secure enclave to protect designs from concept through manufacturing. Next, a dedicated 2.5D GPU enables responsive, high-quality graphical interfaces at the edge, offering realistic visuals at a fraction of the performance and energy cost of traditional 3D processors.

PSOC microcontrollers are fully supported by ModusToolbox, Zephyr, and DEEPCRAFT AI Suite. ModusToolbox features a number of software stacks—including Bluetooth, Wi-Fi, and USB—along with middleware and libraries that can be used to develop custom applications. Zephyr is a small, yet scalable OS with an architecture that allows developers to focus on applications requiring an RTOS.

At Infineon’s OctoberTech 2025 Silicon Valley event held at the Computer History Museum in Mountain View, California, the German chipmaker displayed the company’s PSOC-based edge AI capabilities in applications like advanced sensing. The booth also showcased the analog front-end for a single-chip ECG sensing solution as well as PSOC powering advanced graphics in an AI vision application.

Related Content

• How Edge AI Transforms IIoT and Enables Industry 5.0
• Google Open-Sources NPU IP, Synaptics Implements It
• An edge AI processor’s pivot to the open-source world
• Open-Source ML Tool Aims To Make Edge AI Accessible To All
• Edge AI: The Future of Artificial Intelligence in embedded systems

The post Edge MCUs bolstered by AI design toolchain appeared first on EDN.