Said by: Titanium Media, by pineapple Finance
From February 1 to March 6, in just over a month, ZTE rose 55.6%, Yitong Century rose 74.2%, Wangshu technology rose 93.4%, Huayu Software also rose 43.3%… One common attribute of these companies is that they are all edge computing concept constituents. Edge computing concept has also become A share this wave of gains, the performance of one of the best plate.
Similarly, during this year’s Mobile World Congress in Barcelona, edge computing also made a successful c-bit debut, along with foldable phones and 5G, which became one of the three hottest areas at this year’s Show. In these three hot fields, folding mobile phone is 2C end, is consumer goods; 5G is the communication infrastructure of the new era; As for edge computing, it is not only a new industry and new field, but also the infrastructure and basic capability of the network, computing, storage, application and other near-end overall solutions in the new era.
In the TMT field, there are often a lot of “artificial wind” under the fickleness of the industry and the support of capital, but edge computing is not so, but along with the 5G dongfeng, TMT supply-side reform, and industry digitalization, has become a historical necessity. On the eve of the outbreak of edge computing, the whole industry also vaguely presents three major schools, as well as six characteristics and trends.
Edge computing is not only an industry, but also the infrastructure and basic capabilities of the 5G era
Around this year’s Expo, China Mobile launched the Edge computing “Pioneer 300” Pioneer action, and released the IaaS platform BC-EDGE, Edge Sigma platform, OTII Edge custom server; China Unicom and Wangchu Technology set up a joint venture, Yunjiwisdom, to promote the CDN node sinking and concurrent force edge computing market; Inspur launched the first OTII edge computing server… China’s major mainstream edge computing manufacturers, all through the ba Show to show their muscles.
In fact, The Barcelona show is just a concentrated explosion and release of major manufacturers. As early as around 2000, when content delivery networks (CDN) began to be deployed on a large scale, the embryonic form of edge computing was already formed. CDN is a service that relies on edge server and improves user access response speed through load balancing, network scheduling, content distribution and other functional modules of the central platform. Then in 2003, Akamai and IBM put forward the concept and idea of edge computing.
Akamai was the first to think about edge computing, and in 2016, Internet Technology also started to build edge computing network, and gradually upgraded CDN nodes to edge computing nodes. Akamai also began to make efforts in edge computing at the end of 2018. Li Sheng, regional vice president of Akamai and general manager of Jinan University in China, believes that Akamai is closer to users and has more than 3,000 nodes, so it is more capable in edge computing.
In fact, CDN manufacturers can become the main players in the new field of edge computing, which is not a coincidence, but the “genetic advantage” of CDN manufacturers themselves. Because, with the rapid development of 4K, 8K, VR, Internet of Things, industrial Internet, automatic driving, smart home, intelligent transportation, smart city, AI and other fields, the traditional data center mode has been unable to meet the requirements of timeliness, continuity, security, privacy, intelligence in these fields. Therefore, network, computing, storage, application and other original topological structures must be reconstructed, while edge computing and CDN share the same logic at the bottom of the product.
Like it foundation of Chinese white paper “edge computing – across the traditional data center”, the streaming media, Internet of things, such as artificial intelligence and the explosive development of the upcoming application trend of more immersive, interactive, promoted the focus away from the central data center of data storage and computing, and toward the edge. According to the report “China Manufacturing Internet of Things Market Forecast 2016-2020” released by IDC, 40 percent of data needs to be analyzed, processed and stored at the edge of the network last year alone, and this figure will rise to 50 percent by 2020.
The huge application scenarios and rapid development trend also make the market capacity of edge computing increasing. CB Insights estimates that the global edge computing industry is expected to reach $34 billion by 2023. It is worth mentioning that edge computing not only has vertical industry attributes, but also has horizontal cross-industry “hydropower and coal” attributes. It is the infrastructure and basic capability of 5G, artificial intelligence and Internet of Things in the new era. In the future, industries will embrace edge computing as they did the Internet and mobile Internet.
Industry contention of a hundred schools, three schools of edge computing and three paths
At present, edge computing is still on the eve of the outbreak of the industry. The characteristics and development stage of edge computing determine that it is experiencing a flourishing period of contention and blooming of a hundred flowers. If we comb carefully, we can find that there are at least seven or eight groups in this new field of edge computing, such as cloud computing, CDN, hardware devices, operators, research institutions, industrial alliances, and other small and medium-sized manufacturers. In addition, cloud computing, CDN and hardware equipment have become the three most mainstream sectors. Different schools of origin, not only brings different layout, but also means different evolution path.
First, cloud computing pie. In the face of surging edge computing, traditional cloud computing manufacturers, especially the head of the “3A” manufacturers, on the whole are more open attitude, action is also relatively strong. Amazon’s core product in Edge computing is the AWS Greengrass Edge Computing platform, which has been around since 2016. With this platform, AWS cloud capabilities can be efficiently extended to local devices, and data can be collected and analyzed closer to information sources, while securely communicating with each other over local networks. From the perspective of customer perception, real-time connectivity and response of AWS can be obtained, as well as a secure environment for developing edge applications and analysis.
Microsoft, for its part, came a bit later than Amazon, but in mid-2017 it launched Azure IoT Edge, a service for the Internet of Things that pushes computing power from Azure to Edge devices. It is worth mentioning that Azure IoT Edge also includes the AI module and is fully open sourced via GitHub.
And The domestic Ali Cloud, at the 2018 Cloud Computing Conference · Shenzhen Summit, announced its strategic investment in the field of Edge computing technology, after the launch of IoT Edge computing product Link Edge, also fast. Not only did Intel launch an open framework specifically for Internet of Things applications; It also released edge gateway technology for the Internet of Things; At the same time, open source real-time computing platform Blink.
Besides, CDN school. Compared with the cluster of cloud computing giants, among CDN schools, Akamai is the most important in foreign countries, while Internet hosting technology is the most important in China. Although Akamai began to intensify its efforts to preach edge computing in the Chinese market from the end of 2018, it seems that it has not launched any products or solutions in the real sense, but only emphasizes the virtual level to release its edge computing capabilities.
In contrast, Internet technology seems to be more aggressive. From the beginning of 2018, it began to formally build edge computing platform, provide IaaS and PaaS services of edge platform, and open resources of edge computing platform. At the beginning of the New Year of 2019, the company also joined hands with China Unicom to establish a joint venture company, Cloud Intelligence, which opened the business layout of CDN node sinking and began to make efforts in the edge computing market. Liu Chengyan, chairman of Wangju Technology, believes that it is necessary to aim at the opportunities under 5G scenarios, vigorously layout edge computing, and do a good job in infrastructure construction to provide support for large-scale commercial use of 5G in the future.
Finally, the hardware school. Like the cloud computing group, the hardware group is also a strong cloud, including Intel, Cisco, Huawei, Dell, AMD, ARM and so on. Take Huawei as an example. It has the EC-iot solution for the Internet of Things and the intelligent edge platform IEF for edge computing. Centerm chips that can achieve full scene coverage were announced at the 2018 Fully Connected Conference, and edge computing was also included.
Objectively speaking, in the force edge calculation of the seven or eight schools, each kind of manufacturers have their own advantages, but also have their own challenges. In the competition of edge computing, the key to each manufacturer’s success lies in its understanding of the industry, its speed of turning around, its layout and investment, its ability to form alliances and maintain strategic focus.
There are six characteristics and trends of edge computing
Although edge computing has not broken out in an all-round way, from the current undercurrent, we can already see that edge computing presents six characteristics and trends.
The first is decentralization. In THE TMT industry, especially in the Internet field, “openness” seems to be the first law of the whole industry. And openness often means “decentralization”. When it comes to decentralization, the Internet industry has been working on it for 20 years or so, but there is still a long way to go. At present, shopping, community group buying and social e-commerce have broken the centralization of e-commerce to some extent. We media has cracked the centralization of the media field. Of course, edge computing also breaks down the centralization of computing. From the essence and definition of the industry, edge computing is to distribute network, computing, storage and application from the “center” to the edge, so as to provide intelligent edge services nearby.
Second, non-oligopoly. There has always been a winner-takes-all phenomenon in the TMT industry in many fields. This is true in instant messaging, social, search, security, and more. However, in the field of edge computing, this phenomenon may not exist. The main reason is that the edge of computing is the Internet, mobile Internet, Internet, Internet industry, electronics, IT, cloud computing, AI, hardware equipment, operators, and many other areas of “cross entry”, on the one hand, to participate in all kinds of numerous manufacturers, on the other hand “decentralized” logic underlying the products, is to some extent to the “monopoly”.
Third, the marginalization of everything. As mentioned earlier, edge computing is not only a vertical vertical industry, but also a horizontal infrastructure and basic capabilities. This second attribute will make edge computing as pervasive and universal as IT and the Internet in the early years, cloud computing and mobile Internet today, and artificial intelligence in the future. In the future of the Internet of everything, there are application scenarios for the Internet of everything, and there are application scenarios for edge computing.
Fourth, security. In edge calculation before, most of the user data will be uploaded to the data center, in the process of the upload, the user’s data, especially the private data, such as individual tag data, and bank account password, electricity consumption data business platform, searches, even smart camera and so on, as there is a risk of leakage. Edge computing eliminates this risk at the source because in many cases, data can be processed near the edge instead of being uploaded to the data center.
Fifth, real-time. With the increasing popularity of industrial Internet, autonomous driving, smart home, intelligent transportation, smart city and other scenarios, applications in these scenarios have higher and higher requirements on the speed and efficiency of computing, network transmission, and user interaction. In terms of autonomous driving, for example, it’s almost seconds or even milliseconds. In the face of the massive data created by cameras, radar, laser radar and other sensors in automatic driving, the response, calculation and transmission speed of the traditional data center mode is obviously not enough. At this time, the edge calculation of “near end processing” has naturally become the best choice for the best “real-time” requirements.
Sixth, it is green. This is easy to understand, because the data is processed in the near end, so in the network transmission, central computing, central storage, transmission and other links, can save a lot of server, bandwidth, electricity and physical space and many other costs, so as to achieve low cost, green.
Write at the end:
Edge computing is a long – term, systematic and huge project. It is not only the beneficiary of 4K, 8K, VR, Internet of Things, industrial Internet, autonomous driving, smart home, intelligent transportation, smart city, AI and other developments, but also the catalyst and booster of these fields and scenes. Facing this “new species” of edge computing, it is not a matter of whether to embrace it or not, but how to embrace it. There is no doubt that the first mover will naturally get more technology bonus and time bonus of industrial development.
(This article was first published by Titanium Media)