Welcome Here  Shenzhen Mingjiada Electronics Co., Ltd.

According to market research firm ABI Research, global sales of small and medium model end-side ML devices will reach 4 billion units in 2028, growing at a CAGR of 32 per cent, mainly covering gaming consoles, smart glasses, in-vehicle terminals, security cameras, chatbots, etc. In 2030, 75 per cent of global small and medium model end-side machine learning is based on dedicated hardware. Edge AI has emerged as a key theme for 2025.

As end-side AI applications have well-defined scenarios, prioritise cost optimisation, power consumption and processing power, they offer plenty of room for new market players to grow. Traditional chip majors NVIDIA, Intel, Renesas, NXP and AMD are playing a key role in driving the adoption and implementation of edge AI technologies across all sectors. At the Expo, Renesas unveiled its RZ/V2H-based RZ/V2H - a high-computing power quad-core vision AI microprocessor (MPU) with DRP-AI3 (Dynamically Reconfigurable Processor) accelerator and high-performance real-time processor. Suitable for image processing, dynamic computation required for robotics applications, and applications with complex real-time requirements, it is a microprocessor for applications such as autonomous robots and machine vision in factory automation.

In September, NXP Semiconductors announced the launch of its new i.MX RT700 crossover MCU family, designed to empower smart AI-enabled edge-end devices such as wearables, consumer medical devices, smart home devices and HMI platforms.

Torchcore Technology launched a new generation of MMSCIM end-side AI audio chips based on a CPU+DSP+NPU triple-core heterogeneous core architecture, including ATS323X, ATS286X and ATS362X.Overseas emerging supplier of AIoT chips, Aon Devices, launched a chipset targeting natural language processing, chatbots, and digital assistants, Halio Technologies has a rich portfolio of edge AI processors, the Hailo-15 series for video and image processing, and the Hailo-8 AI accelerator.