AI evaluation kit targets inference and always-on workloads on PSOC™ Edge E84
The KIT_PSE84_AI evaluation kit is intended for assessment of artificial intelligence and machine learning workloads on the Infineon PSOC™ Edge E84 microcontroller. The board is configured to demonstrate the interaction between the high-performance compute domain, based on the Arm® Cortex®-M55 with Ethos™-U55 NPU, and the low-power always-on domain, which combines a Cortex-M33 core with Infineon’s NNLite neural network accelerator.
The platform supports evaluation of common edge AI use cases such as acoustic activity detection, voice processing, and sensor-based inference. Audio interfaces, sensor connectivity, including 60GHz radar and pressure sensors as well as memory access are provided to allow execution of neural network models across shared SRAM and non-volatile RRAM resources. This enables analysis of task partitioning, power-domain interaction, and inference latency under representative operating conditions.
The kit is supported by the ModusToolbox™ development environment and integrates with DEEPCRAFT™ Studio tooling for model deployment and validation. It is intended for engineers evaluating AI-enabled embedded architectures where low-power, always-on inference and higher-performance processing must operate within a single device.
