The HL-8 Series is a high-performance AI accelerator powered by Hailo-8, delivering up to 26 TOPS of AI inference performance without the need for external DRAM, thanks to its on-chip memory–based architecture.

By adopting a dedicated architecture optimised for neural network processing, the HL-8 Series achieves both high computational efficiency and low power consumption. This approach reduces overall system cost and simplifies hardware design, significantly improving reliability in edge AI environments.

Designed for seamless integration with the XR Series and NX Series, the HL-8 Series can be deployed across a wide range of AI-driven applications, including medical image analysis, industrial machine vision, and intelligent automation. By enabling local AI inference, it reduces reliance on cloud infrastructure, minimises latency, and provides an optimal solution for edge AI environments where data privacy is critical.