WAIC 2026 On the Scene | Xiangcheng Electronics Unveils a Cutting-Edge Domestic Edge AI Solution
The 2026 World Artificial Intelligence Conference (WAIC) officially opened in Pudong, Shanghai, with this year’s event focusing on… “Smart Partner, Co-Creating the Future” Centered around a core theme, the event brings together over 1,100 AI‑related upstream and downstream enterprises, focusing on four key areas—domestic foundational computing power, edge‑based vision, embodied intelligence, and industrial AI—thus serving as a pivotal industry platform for monitoring the practical deployment of domestic AI solutions and the iterative advancement of domestically produced hardware.

As a company that has been deeply engaged in embedded edge AI hardware for over a decade, Xiangcheng Electronics brings with it… XC3588E Smart Motherboard + DEEPX DX-M1 AI Acceleration Card A comprehensive integrated solution makes a major appearance, with “Applications of Smart Motherboards and Computing Modules in Artificial Intelligence” As the exhibition theme, we conducted a live demonstration for the first time of a dual‑compute‑power hardware platform optimized for China’s Kylin operating system, fully showcasing two core edge‑side AI vision capabilities—multi‑object recognition and real‑time action‑tracking and recognition—providing an autonomous, controllable, low‑power, high‑real‑time edge AI hardware foundation for industries across the board.

Self-developed XC3588E smart motherboard: the core platform for domestically produced computing power.
The XC3588E is Xiangcheng Electronics’ self-developed flagship RK3588 smart motherboard, powered by a domestically produced Rockchip 8nm octa-core 64-bit processor (4×A76 + 4×A55), with built-in… 6 TOPS native NPU It can independently execute lightweight AI inference tasks, eliminating reliance on cloud‑based computing power. At the hardware level, it boasts exceptional industrial‑grade compatibility: equipped with nine USB ports, six serial interfaces, dual Gigabit Ethernet ports, dual CAN buses, and multiple GPIOs; it supports a full range of display interfaces including EDP, MIPI, HDMI, and LVDS; and it features a standard M.2 M‑KEY slot for seamless integration of the DEEPX DX‑M1 accelerator card, enabling elastic scaling of computational capacity. Fully compatible with the domestically developed Kylin operating system. , synchronous compatibility Multi-system support for Android and Ubuntu , significantly lowering the development barriers for domestic‑solution projects and truly achieving full‑stack, independently controllable capabilities with “domestic chips + domestic operating systems.”

DEEPX DX-M1 AI Acceleration Card: Elastic Scaling for High-Density Inference Computing Power
The DEEPX DX-M1 AI accelerator card, designed to work with the XC3588E motherboard, serves as the core of the entire solution’s computational performance boost: a single card can deliver up to 25 TOPS INT8 inference computing power , with the overall system power consumption kept below 3W, enabling high‑density, low‑power computing scalability; featuring an M.2 plug‑and‑play design that requires no complex driver modifications and is compatible out of the box with the XC3588E motherboard. When combined, the resulting computational performance sees a substantial leap, providing stable support. Synchronous real-time analysis of 1080P video streams from Route 8 and above. , perfectly meets Multi-object recognition, human pose tracking and recognition Supports high‑load scenarios; the software ecosystem is fully compatible with mainstream deep learning frameworks such as PyTorch, TensorFlow, and ONNX, enabling one‑click model quantization and deployment, thereby significantly shortening the time-to-market for customers’ AI projects.

On-site live demonstration: object recognition + action‑tracking recognition, vividly showcasing the power of edge‑side AI.
At the Xiangcheng Electronics booth, the entire set… XC3588E Smart Motherboard + DEEPX DX-M1 Acceleration Card + Domestic Kylin Operating System The solution deploys a real‑world demonstration terminal to showcase, in real time to attendees, the capabilities of two core computer vision algorithms: 1. Object recognition : Synchronously connects to cameras, enabling millisecond‑level object detection, classification, and coordinate localization for workpieces, personnel, shelf‑stocked items, and roadblock obstacles, with smooth, lag‑free video and instant recognition. It is well suited for industrial quality inspection, smart retail, in‑vehicle edge‑AI applications, and other scenarios.
2. Human Motion Tracking and Recognition : It captures human body joint positions in real time, accurately detecting violations, falls, boundary crossings, and pose‑based interactions, and can be directly deployed in scenarios such as construction site safety management, warehouse AMR‑human collaboration, service robot interaction, and AI digital‑human terminals.
Building a new future for edge AI, powered by domestically developed hardware.
Looking ahead, Xiangcheng Electronics will continue to deepen its R&D in embedded edge AI, iteratively refine its flexible computing‑power motherboard and accelerator‑card solutions, and expand compatibility with an even broader range of domestically developed software and hardware. By providing a reliable, widely accessible domestic edge‑hardware platform, the company will support the digital transformation of industries across the board—ranging from industrial automation and intelligent robotics to visual monitoring—and work hand in hand with ecosystem partners to co‑create a new domestic AI‑hardware ecosystem.
Recommended
From data center infrastructure monitoring to intelligent manufacturing, from energy power plants to AI‑driven security, in the era of domestically developed edge computing, Xiangcheng Electronics’ industrial control host series provides a stable and reliable hardware foundation, helping industries across the board achieve digital and intelligent transformation.
The XC3576H industrial control motherboard has now been fully adapted to Ubuntu 26.04 LTS, offering developers a high-performance, long‑support‑cycle, and feature‑rich ARM‑based open‑source solution for industrial applications.
The IPCA_3588E industrial control host features a self-developed XC3588E industrial motherboard and is built with exclusively new, original‑brand chips, delivering an integrated edge computing terminal that combines domestically produced chips with a domestic operating system.
