1. With the rapid growth of global data, centralized cloud computing has been unable to fully cope with it, and just-needed scenarios have emerged. At present, the number of Internet of Things connections in China will increase from 5.5 billion in 20 19 to14.8 billion in 2023, with a compound annual growth rate of 2.8 1%. The amount of data perceived by the Internet of Things has surged, and the data types have become more and more complex and diverse. IDC predicts that by 2025, the amount of data generated in China will increase by 48.6ZB every year.
2. Chip: FPGA meets the requirements of performance, power consumption and delay at the same time. Different from centralized cloud computing, the physical environment of edge cloud computing is complex and diverse, and space, temperature and power system are often not the best state. But at the same time, the edge requires extremely high real-time performance and computing performance, and the traditional CPU architecture is difficult to meet the needs of the edge cloud. Intel, Xilinx and other international chip giants are constantly adding more FPGA chips and launching hardware platforms supporting CPU+FPGA heterogeneous computing. The prosperity of the underlying chip industry will support the application of edge cloud computing in various fields, and generate will continue to show new vitality.
3.5G technology upgrade, Wi-Fi complements indoor scenes. According to the data of the Ministry of Industry and Information Technology, by 2020, China will open more than 710.8 million 5G base stations, realizing the coverage of cities above prefecture level and key counties and cities. It is expected that edge cloud computing will also land in stages with the popularization of applications in the 5G industry. In addition, Wi-Fi technology is also developing in the direction of higher throughput, larger coverage area and lower delay. The advantages of Wi-Fi in indoor scenes make it an important supplement to 5G, both of which will contribute to edge cloud applications.
4. Cloud computing: the cloud on the enterprise is normalized, and the cloud sinks natively to realize the integration of the cloud edge. In recent years, the popularity of cloud locality has continued to rise. Cloud native technologies and concepts, including containers, microservices, DevOps, etc. It emphasizes loosely coupled architecture and simple and convenient scalability, aiming at achieving a consistent cloud computing experience on different infrastructures through unified standards. Compared with virtual hosts, cloud nativity is more suitable for the scenario of edge cloud computing, which can provide integrated application distribution and collaborative management for the cloud side and solve the problems of large-scale application delivery, operation and maintenance, management and control on the edge side.
5. "New infrastructure" is overweight, and benchmarking applications such as industrial internet lead industrial integration. "New infrastructure" is the key direction of the 14 th Five-Year Plan. By optimizing the resource structure of computing power, high-frequency calling and low-delay service requirements are allocated to the edge data center, which promotes the construction of edge networking of 5G bearer network and creates conditions for computing power and network sinking to the edge. At the same time, industrial policies such as industrial internet, car networking and telemedicine explicitly mention edge computing, promote key technology research, standard system construction and software and hardware product research and development, and promote the integrated application of edge cloud in typical industries.
Application scenario
1. video acceleration and AR/VR rendering
Intelligent video acceleration based on mobile edge computing can improve the low efficiency of mobile content distribution: Radio Analyticsapplication is deployed in the mobile edge computing server of wireless access network to provide the video server with real-time throughput index of wireless downlink interface, thus helping the video server to make more scientific TCP (Transmission Control Protocol) congestion control decisions and ensuring that the application layer coding can match the estimated capacity of wireless downlink. In addition, because the AR/VR information (user position and camera angle) is highly localized, it is better to process these information locally (mobile edge computing server) rather than in the cloud, so as to minimize the AR delay and improve the accuracy of data processing.
2. Internet of Vehicles (Intelligent Transportation)
After applying the mobile edge computing technology to the Internet of Vehicles, the cloud of the Internet of Vehicles can sink to a highly distributed mobile communication base station. The mobile edge computing application directly receives the localized data from the in-vehicle application (APP) and the road sensor in real time, then analyzes it, and transmits the conclusion (danger warning information) to other networked vehicles in the nearby area with extremely low delay. The whole process can be completed in milliseconds, so that the driver can make a decision in time.
3. Industrial Internet
Edge computing has always been closely related to industrial control systems. The industrial control system with industrial Internet interface is essentially an edge computing device, which solves the contradiction between the high real-time requirement of industrial control and the uncertainty of Internet service quality. In the infrastructure layer, field devices are connected to the industrial data platform in a flat interconnection way through industrial wireless and wired networks; In the data platform, according to the process and process model of the production line, the field equipment is dynamically managed and combined through service composition, and connected with MES and other systems. Industrial CPS system can support the flexible adaptation of production plan to the change of production line resources, as well as the rapid replacement of old manufacturing equipment and the online introduction of new equipment.
4. Internet of Things gateway services
By adopting edge computing technology, the edge computing convergence node is deployed near the terminal equipment of the Internet of Things to provide sensor data analysis and low-delay response. Among them, the computing power and storage capacity of the edge computing server can provide services for the following five aspects: business aggregation and distribution; Analysis of device messages; Decision logic based on the above analysis results; Database login; Remote control and access control of terminal equipment.
traditional Chinese scale of weights
It is estimated that the scale will exceed 50 billion yuan in 2025, with a compound annual growth rate of 43.3%. According to the survey data of ICT Institute in May 2020, less than 5% of enterprises in China use edge computing, but the planned proportion is as high as 44.2%. It can be seen that although edge cloud computing is still in the embryonic stage of development, the future growth space is very broad. According to the calculation of iResearch, the market scale of edge cloud computing in China will be 96,543.8 billion yuan in 2020, of which the market scale of regional, on-site and Internet of Things edge cloud will reach 3.7 billion yuan, 3.8 billion yuan and1600 million yuan respectively. It is estimated that by 2025, the overall edge cloud scale will increase to 55 billion yuan, with a compound annual growth rate of 44.0%, among which the regional edge cloud will lead the growth rate with mature scenarios such as interactive live broadcast, vCDN and car networking. In 2030, the market scale of edge cloud computing in China is expected to reach nearly 250 billion yuan, and the compound annual growth rate from 2025 to 2030 is lower than the previous five years. During this period, the industrial Internet, smart parks, smart logistics and other scenes of the edge cloud will mature rapidly.
Related listed companies
ZTE Corporation (abbreviation of ZTE Telecom Equipment Corporation)
ZTE provides operators with a full-scenario MEC solution, which breaks the traditional closed telecommunications network architecture, deeply integrates the mobile access network with the Internet, and meets the individual needs of customers at the edge of the network. ZTE's general edge computing solution includes MEP capability open platform, lightweight edge cloud, a full range of edge-oriented servers and edge acceleration hardware, providing various hardware options such as general hardware and special integrated hardware, and deeply integrating OpenStack and Kubernetes to provide a unified edge cloud management system for MEC applications, which is convenient for operators to deploy MEC according to local conditions.
Wang Su technology
The company's edge computing platform is based on four platforms: cloud host, container, function computing and network. Basic services such as edge cloud host, edge cloud container, edge cloud function, SD-WAN and edge cloud security are deployed on edge computing nodes, as well as various internal and external application modules. Combined with the customer's business scenarios and needs, the company tried to integrate the output solutions.
Chuling information
With the rapid development of 5G and AI technologies, the company continues to build three core capabilities of edge computing represented by fixed-mobile intelligent connection (5G+ fixed)+data processing (DPI)+AI. Over the years, the company has deeply cultivated technologies such as enterprise (industry) intelligent connection network, vertical industry edge application DPI (security, Internet of Things), video and other industries (enterprise) intelligent applications, and initially formed a "cloud edge" collaborative edge computing ecology. On the market side, the company not only focuses on the traditional operator market, but also actively expands the government and enterprise industries and large and medium-sized enterprise markets, and has won several bids. In the third quarter, the company demonstrated its cooperation with China Unicom in edge computing, and carried out the carrying capacity test of CUNOS in 5G environment.
Quoted content
1. Research Report "China Edge Cloud Computing Industry Outlook Report"
2. Research report "Edge Computing: an important part of computing network, and the industry is in the ascendant"
Risk warning
1. The underlying related technologies are developing slowly, and the demand for edge computing is less than expected.
2.5G progress is not up to expectations.