According to the Wise News APP, since Chat GPT became popular, with the hundredfold computing power demand explosion under the current scenarios including large-scale chat AI, medical care, smart factories, smart roads and so on, the current computing power development mode mainly based on traditional computing clusters has been unable to meet future needs. A computing power revolution of "from central to ubiquitous", "with energy consumption as the boundary", and "cloud-edge-end integration" is imminent.

Compared to cloud computing, edge computing refers to the practice of putting more computing tasks at the network edge to reduce latency, avoid bandwidth occupation caused by a large amount of data transmission, and protect local sensitive data security among other unique advantages, rather than centralizing all computing resources in the cloud as in the traditional cloud computing architecture.

With the popularity of Artificial Intelligence (AI), the approach of "training in the cloud and reasoning on the edge" has been widely recognized. By deploying trained machine learning models in edge devices, edge devices can perform AI inference tasks quickly and efficiently, accelerating the landing of more and more AI applications.

On May 29, according to the Hong Kong Exchanges and Clearing Limited (HKEX) official website, CloudWalk Technology Co., Ltd. (CloudWalk) went through the HKEX listing hearing, with sole sponsor being PU YUAN International FinTech Group Limited.

CloudWalk is a service provider that develops, operates, and provides IDC solutions and edge computing AI services.

In 2022, CloudWalk launched the Lingjingyun brand to provide infrastructure and computing services that are different from IDC solution services, combining content distribution networks with other functions provided by hardware-centric devices, allowing customers to build edge computing infrastructure as the basis and provide digital experience ("edge computing services") at the Internet edge. It has been selected in the "2023 Edge Computing Industry Atlas","Collection of Innovation and Application Cases of the Industry Metaverse" at the World Artificial Intelligence Conference, and has been selected as one of the "Top 20 Chinese Edge Computing" in 2023 and 2024 for two consecutive years.

Its EdgeAIoT edge intelligent interconnection solution transfers cloud computing AI applications, networks, storage, computation, and other capabilities to the edge, and extends public cloud capabilities to the end close to the terminal device, providing customers with low-latency, high-performance, secure, and reliable intelligent edge computing services through flexible and configurable diverse AI algorithm libraries, edge computing intelligent management platforms, and edge computing hosts with super computing performance, large capacity storage, small and flexible configuration.

For example, take the currently popular autonomous driving technology. Autonomous driving systems are extremely complex, with many closely integrated technologies, including sensing, positioning, perception, decision-making, and smooth interaction with cloud platforms for generating high-definition maps and storing data. These complexities pose many challenges for designing autonomous driving edge computing systems.

The VEC system for in-car-edge computing needs to process a large amount of data in real-time. Mobile VEC systems usually have very strict energy consumption restrictions. Therefore, adequate computing power must be supported by reasonable energy consumption to ensure the safe operation of autonomous driving vehicles even when traveling at high speeds.

When designing an edge computing ecosystem for autonomous driving vehicles, the primary challenge is how to achieve real-time processing of data, providing sufficient computing power, ensuring reliability, scalability, security, and cost control while ensuring the safety of autonomous driving vehicle users, and enhancing user experience.

(Information source: GTI official website, Yicun Capital)

The fusion of edge computing and AI has enormous potential to change global industries.

By using localized processing, real-time insight, and optimized resource utilization, the full potential of AI can be fully released while protecting sensitive data and promoting organizational entry into the AI era.

In this regard, global technology giants have begun to shift their focus more towards the field of edge computing AI.

For example, Huawei, Qualcomm, and NVIDIA have all launched edge AI products. In March of last year, Huawei released the Smart Search function in July last year, which miniaturizes the model; Qualcomm China demonstrated a Stable Diffusion with model parameters exceeding one billion for the first time on an Android phone. Huawei and Qualcomm have to some extent verified the feasibility of high-performance edge AI, and it is expected that AI large models can be experienced at the edge through model compression and network intelligence.

In addition, NVIDIA's Jetson platform, a series of modules and development tools designed specifically for edge computing, can handle complex data processing tasks at the edge, combined with NVIDIA's GPU technology and deep learning algorithms, and achieve efficient image and video analysis, object recognition, and voice processing.

The entry of technology giants such as Huawei, Qualcomm, Nvidia, etc. is driving the increasing maturity of the overall industry and ecosystem of edge computing and AI, making edge computing AI an important support for improving the level of information development and creating new digital economic momentum.