The Hailo-8 M.2 AI Acceleration Module HM218B1C2YAE is a compact edge AI accelerator starter kit designed to bring high-performance neural network inference to compatible embedded and edge computing platforms.
Powered by the Hailo-8 AI processor, the module delivers up to 26 TOPS of AI processing performance. This enables demanding deep learning and computer vision workloads to run directly on edge devices, reducing dependence on cloud processing and helping achieve lower latency for real-time applications.
The HM218B1C2YAE variant uses the compact M.2 2230 A+E Key form factor. Its small footprint makes it suitable for embedded computers, industrial PCs, edge gateways, smart cameras, and compact AI-enabled systems where available board space is limited.
Communication with the host platform is provided through a PCIe Gen 3 x2 interface. The module supports both x86 and ARM host architectures, making it suitable for integration into a wide range of embedded computing platforms.
The Hailo-8 platform is designed for real-time, low-latency AI inference and supports multi-stream and multi-model processing. It can therefore be used in applications involving object detection, image classification, intelligent video analytics, automated inspection and multi-camera computer vision.
Hailo provides a complete AI software ecosystem for developing and deploying models on the accelerator. Supported frameworks include TensorFlow, TensorFlow Lite, Keras, PyTorch and ONNX, together with HailoRT, the Hailo Dataflow Compiler and model development resources.
The HM218B1C2YAE is specified for an extended operating temperature range of -40°C to +85°C, making it suitable for industrial automation, transportation, outdoor systems and other demanding edge environments.
The starter kit also includes thermal-management accessories for easier evaluation and integration of the module into compatible systems.
