Infineon launches DEEPCRAFT™ AI Suite to accelerate Edge AI innovation

Infineon Technologies AG has expanded its edge artificial intelligence portfolio with the launch of the DEEPCRAFT™ AI Suite, a new collection of software, tools, and ready-made solutions that help developers integrate AI capabilities directly into edge devices.

The suite is designed to work seamlessly with Infineon’s recently introduced PSOC™ Edge microcontrollers, combining low power, high performance, and robust security for next-generation embedded applications.

According to Infineon, DEEPCRAFT is built to simplify and speed up the development of voice, audio, radar, time-series, and computer-vision applications on embedded systems. It includes DEEPCRAFT Studio, an integrated environment where developers can train, optimize, and deploy models tailored for Infineon hardware.

The platform will support popular machine learning frameworks such as PyTorch, TensorFlow Lite, and Keras, and provide a Model Converter tool to streamline model migration and optimization.

The new suite also introduces an AI Hub containing more than 50 pre-built resources, including open-source and proprietary models, development tools, and reference projects. These resources are aimed at helping engineers reduce development time and accelerate design cycles for industrial, consumer, and automotive edge AI systems.

One of the DEEPCRAFT Suite’s standout features will be its ability to deliver AI at extremely low power, enabling always-on voice and sensor monitoring while consuming less than 1 milliwatt (mW) of power.

This ultra-efficient operation significantly extends the battery life of portable and battery-powered devices, allowing smart wearables, wireless earbuds, and home automation systems to perform AI inference locally without frequent recharging. Infineon emphasizes that this power efficiency is achieved without compromising responsiveness or model accuracy.

The supporting hardware, PSOC Edge, integrates Arm Cortex-M cores with built-in machine learning acceleration and Infineon’s Edge Protect Category 4 (EPC4) security framework. It also carries PSA Certified Level 2 and Level 4 compliance, ensuring secure model execution and data protection across devices.