The world's edge computing market will reach 250 i

  • Detail

In 2024, the global marginal computing market will reach US $250billion, or it will breed Unicorn enterprises. Every new information technology, from birth to maturity, seems to be crossing the river by feeling the stone, but in fact, there are internal laws to follow. As the edge computing bridging 5g, artificial intelligence, cloud computing and IOT, it is the leader in the emerging technology trend in 2020. According to IDC analysis, more than 50% of the data in the future will depend on the maturity of edge computing for analysis and calculation. It is expected that the global edge computing market will reach 250billion US dollars in 2024. After the initial exploration, the promising edge computing gradually sees the future development path. At the edge computing industry summit held recently, the edge computing industry alliance (hereinafter referred to as ECC) once again knocked the blackboard for the development of edge cloud collaboration, and promoted the cooperation between industry and academia on the theoretical innovation of edge cloud collaboration, paving the way for industrial prosperity and application implementation

from concept to practice

edge cloud collaboration has entered the era of 2.0

edge computing is booming because it meets the needs of providing edge intelligent services nearby, and it is also promoting the computing model from centralized cloud computing to more distributed edge computing. Cloud and edge have their own strengths. Only when edge computing and cloud computing work closely together can they better support various businesses, and the value of edge computing and cloud computing will be further amplified

in 2018, we launched the white paper on the synergy between edge computing and cloud computing, which played a very good role in building consensus for the entire industry. Huang Huanqing, chairman of the ECC demand and overall group, said in an interview with China electronic news that almost every cloud enterprise is now proposing cloud side collaboration. This concept has been deeply rooted in the hearts of the people, but the understanding of how to implement it and how to apply new technologies is still vague. This year, we launched the white paper on cloud side collaboration 2.0 technology, hoping to promote the implementation of cloud side collaboration on the basis of consensus. Among them, we have put forward some reference technical architectures and introduced some excellent technologies

Ruan Binfeng, project manager of ECC edge cloud collaboration 2.0 industry, said: we have jointly discussed with 15 partners, such as Tencent, Unicom, Haier, etc. From the perspective of 2.0, we believe that edge cloud collaboration 2.0 has three main collaboration directions, namely, application collaboration, service collaboration and resource collaboration. From these three levels, each layer has different challenges

the goal of application collaboration is to register access at one point of edge applications and obtain a consistent experience at other distributed edge application access points. For the implementation practice of edge computing, application collaboration is the core of the whole system, involving all aspects of cloud, edge, management and end

in application collaboration, we can see many challenges. The first challenge is how to manage large-scale edge nodes in a relatively centralized cloud, which is actually a great challenge. At present, the industry has also proposed many technologies to solve such problems, such as P2P technology, buffer technology, etc., which we also mentioned in the white paper. Ruan Binfeng said that the second challenge is that there are applications on the cloud and on the edge. For example, smart cities and smart transportation are typical scenarios. How to achieve the collaboration between cloud and edge applications, and how to achieve unified management and unified distribution. The third challenge is the stability of operation. There will be many problems in the edge scene. The network and edge facilities are unstable. In this case, how to ensure the continuity and reliability of edge Cloud Applications. But I want to achieve 200kg. People are also discussing whether we should shield the differences of underlying infrastructure for application developers as much as possible. They just need to pay attention to their own application development

the goal of service collaboration is to improve the efficiency of the formation of edge applications. Service collaboration mainly includes two aspects: on the one hand, the capabilities provided by cloud services from the central cloud and cloud ecological partners, including the capabilities of data, intelligence and application enablers; On the other hand, the cloud native architecture provides a complete set of processes for the access, discovery, use, operation and maintenance of edge services

now we can see that the specifications of various services provided by different enterprises on the cloud, such as AI services, data services and application services, are different, which will make various industries fragmented. We hope to promote the standardization of services and promote the standard access of services through the industry's standard framework. Ruan Binfeng said that a large part of service collaboration is intelligent services. Intelligent services require a large number of samples. There are challenges in how to improve incremental knowledge, realize collaborative reasoning, ensure security and privacy, and how to store a large amount of data obtained in edge computing

realizing resource collaboration can not only simplify the difficulty of application development, but also improve the efficiency of edge resources. For resource collaboration, we can apply some ideas and ideas of cloud computing. However, there will be more connotation expansion in edge computing. In addition to the collaboration of how to select fixtures for computing electronic universal testing machine and storage network, it also involves the collaboration of traffic. For example, how to do global scheduling, how to speed up the whole domain and browsing, and how to do the underlying computing and storage network resources, we can consider the problem of resource collaboration from these three dimensions. Ruan Binfeng said

from industry to academia

edge computing calls for theoretical innovation

the development of edge computing to the present will certainly produce a new technical system, just as the development of cloud computing has also produced a new technical system, so it needs the cooperation of academia and industry to overcome the problems encountered in the process of industrial practice. Huang Huanqing said

at the edge computing industry summit, Yao Jianquan, academician of the Chinese Academy of Sciences and chairman of the smart city industry alliance, said in December 2016 that edge computing is an inevitable trend of the continuous evolution of the computing architecture of the digital world, and the cloud, edge, and end collaborative computing architecture is the same as the human brain, spinal cord, and peripheral nervous system architecture. In terms of sociological development, the evolution process from unit work to simple separate collaboration, and then to the formation of a huge social collaboration organization is the same as that of cloud, edge, and end computing

world digital intelligence has evolved from progressive monomer computing to distributed network computing, and is evolving towards heterogeneous, collaborative, and comprehensive ubiquitous intelligent computing. Yao Jianquan said that edge computing should consider the following theoretical issues. First, the computability theory should consider the possibility of different computing modes and multi-mode collaborative computing according to the characteristics of computing tasks; The second is fuzzy computing theory, which needs to consider the scale, density, fuzzy objectives and preset conditions of fuzzy computing; Third, complexity theory, which needs to take into account the reliability of computing tasks and the overall project in complex scenarios such as multi consumption, multi constraints, heterogeneous dynamics, etc; The fourth is the efficiency theory, which needs to be optimized and improved from the perspective of efficient data cache, network collaborative transmission, elastic training reasoning and so on

Yao Jianquan also pointed out that it is necessary to continue to explore supporting theories in terms of edge computing infrastructure, platform architecture, and application services. In terms of infrastructure, how to achieve scalability, heterogeneity and collaboration of computing; In terms of platform architecture, how to have high availability, flexibility and universality; In terms of application services, how to divide the resources of services, adjust the operation and maintenance capabilities, and whether the services have self adaptability, isolation and immunity. In a word, these problems are still immature. I hope that our academic and business circles can sit down together and do some theoretical research to guide the development of the entire edge technology and the progress of edge industries. Yao Jianquan said

after edge computing, we have a new opportunity to repudiate this over centralized cloud service model and spread the data. Guo Deke, deputy director of CCF distributed computing and systems special committee, said in an interview with China electronic news that there are theories and methods of distributed computing in cloud computing, and edge computing is more decentralized than distributed cloud computing. We must consider how to use the theories and methods of distributed computing in such a decentralized situation to meet the needs of users while having the convenience and quickness provided by edge computing, Ensure the consistency and reliability of the experience and the overall performance of cloud side collaboration

from centralized to decentralized

the new computing pattern will breed Unicorn enterprises

at present, the cloud computing market has been highly monopolized, and some people say that interconnection will only need six cloud computing enterprises in the future. Guo Deke told that the living space of small cloud enterprises will be smaller and smaller

under the trend of cloud side collaboration, will edge computing in the future also become a feast for giants? Is there room for the growth of emerging enterprises

in the field of edge computing, it must be very important, such as adhesives, catalysts and internal release agents. There is no doubt that Unicorn enterprises will emerge. Huang Huanqing believes that the reason for this is that we can refer to the development of cloud services. From the emergence of the cloud concept to the real start of industrialization, and finally to the emergence of service models and business models that have a great impact on social economy, the whole experience takes a very long time. Objectively speaking, the big Unicorn companies generated around the cloud, or the companies that have achieved commercial success, are things of the past five years. The real rise of edge computing has also begun in 2015 and 2016. From the perspective of the development cycle of edge computing, it is only at the stage when there is a need for cooperation between industry and academia

this kind of cooperation will bring new technology system, which is similar to the new technology system produced by cloud. It will certainly be produced, but it is not clear what it is. Once the new technology system is produced, it will feed back to the industry, which requires a cycle to achieve. Huang Huanqing said, so I think Unicorn enterprises will certainly emerge, but because the whole industry is not yet mature, many new things are still being explored

Ruan Binfeng said that from several aspects of edge computing, it is difficult for emerging enterprises to expand the platform layer, but the application layer of edge computing covers a lot of scenarios, in which there will be a lot of opportunities

ecc has always said that edge computing has six value industries, including industrial manufacturing, digital energy, intelligent transportation, smart city, etc. We believe that in the next 2-3 years or even 3-5 years, marginal computing applications in these industries will rise fastest. Huang Huanqing said that with the development of 5g, the construction of operator mec (multi type edge computing) has actually entered the stage of commercial deployment, and is also accelerating. Moreover, around personal oriented applications such as cloud games, cloud, AR, VR and live broadcasting, enterprises in Unicorn form may emerge faster in these fields that will explode in the short term

Copyright © 2011 JIN SHI