Miao Xinjun's team summarized and answered three questions about the computer industry concerned by the current market:
(1) application scenario of digital RMB
(2) comparison between AI algorithm companies and traditional industry service providers
(3) how to view the prosperity of industrial software industry in the future?
application scenario of digital RMB
the digital RMB pilot has rich scenes, covering retail trading, government services, securities trading and other fields. {23456} as of October 22, 2021, China has more than 3.5 million digital RMB pilot scenarios ; In 2022, the pilot of digital RMB in the Winter Olympic Games scene was steadily promoted, and three large-scale digital RMB activities were carried out, with 403000 landing scenes. After the official promotion, the application scenario of digital RMB is expected to be further expanded.
among many pilot scenarios, we believe that smart contract is expected to become one of the most widely used scenarios. smart contract is a digital protocol that uses algorithms to write contract treaties and deploy them on the blockchain to execute automatically according to rules. By loading smart contracts, digital RMB can realize conditional payment under specific conditions such as trigger time, scene and role, and can easily solve the problems of tripartite fund trust, targeted loan consumption, targeted use allocation and so on. At present, the cloud Side Smart contract has been implemented in the smart leasing scenario jointly created by ABC Shenzhen Branch and Huawei, and xiong'an new area is also piloting the settlement of photovoltaic power generation by means of digital RMB + smart contract; In the future, digital RMB loaded with smart contracts is also expected to become an intelligent pricing and payment tool in carbon emission trading, stock futures trading and other markets.
comparison between AI algorithm company and traditional industry service providers
from the perspective of AI industry solution architecture, AI algorithm companies and traditional industry service providers are located in different architecture layers, and AI algorithm companies are mainly in the "cloud PAAS" layer.
AI solution architecture can be roughly divided into five layers:
1) participants in the side and end layer mainly include NVIDIA and Huawei in the direction of edge reasoning chip, and Haikang, Dahua and Yushi in the direction of camera edge computing.
2) IAAs layer participants include Inspur, Huawei, dawn, Dell and Xinhua in the server direction; In the direction of public cloud, there are large public cloud manufacturers such as Alibaba, Huawei, Baidu and Tencent; In the direction of storage service and network service, there are Haikang, Dahua and Yushi.
3) the PAAS layer can be divided into three sub layers . There are Beijing Haitian Ruisheng Science Technology Ltd(688787) , softpower and other participants in the data acquisition, cleaning and labeling layers; In the training framework layer, there are participants such as Google and Facebook; In the algorithm layer, AI algorithm companies such as Shangtang, Kuangshi, the fourth paradigm, Yuncong, hongruan and Yitu participate.
from the perspective of the advantages of each company in the AI industry chain, each company has different comparative advantages. It is important to find a suitable landing scenario , and consider which kind of enterprise can achieve the comprehensive optimization (operating cost + sales cost + R & D cost) in this scenario. In some large scenarios, the participants in the industrial chain mainly focus on cooperation; In some small scenarios, the participants may have fierce competition.
The comparative advantage of hardware / cloud service provider lies in the low operating cost brought by the scale effect; The comparative advantage of scenario owner lies in its deep understanding of the business, low delivery cost and low R & D cost caused by the direct accumulation of scenario data; The comparative advantages of the original industry IT service provider lie in the deep understanding of the business, the low delivery cost and low sales cost brought by the reuse of the original sales system. The comparative advantage of AI algorithm company lies in the low R & D cost (i.e. high R & D efficiency) under the condition of high availability of the algorithm.
to sum up, we believe that AI algorithm company has comparative advantages in scenarios requiring high calculation accuracy and speed; Haikang, Dahua and other companies have comparative advantages in scenarios with high operating costs / sales expenses and relatively low requirements for algorithm accuracy and speed.
how to view the prosperity of industrial software industry in the future?
we believe that judging from the policy and long-term demand of the industrial software industry, the current PMI continues to rebound. Under the dual carbon background, the digitization process of high energy consuming industries is accelerated, and the long-term prosperity trend of the industry is clear.
Requirements:
1) the PMI value continues to rebound, the high-tech and equipment manufacturing industry continues to expand, and the long-term digital transformation trend remains unchanged. In December, the PMI value was 50.3%, up 0.2 percentage points from the previous month, higher than the critical point, and the prosperity level of the manufacturing industry continued to rise. we believe that PMI reached the bottom in October of 21 and continued to rise for two consecutive months in November and December, indicating that the downstream boom has warmed up, manufacturing production has maintained expansion and accelerated digital transformation expenditure. At the same time, the long-term trend of digital transformation is determined. In the long run, digital transformation helps to improve enterprise efficiency and "quality" competitiveness.
2) the supply side reform of high energy consuming industries has begun, and efficiency is the core measurement basis. on November 15, 2021, in order to implement the several opinions on strengthening energy efficiency constraints and promoting energy conservation and carbon reduction in key areas, the national development and Reform Commission and other departments issued the benchmark level and benchmark level of energy efficiency in key areas of high energy consuming industries (2021 version) to guide all localities to scientifically and orderly carry out the technical transformation of energy conservation and carbon reduction in high energy consuming industries. In particular, it is emphasized that all localities should clarify the transformation, upgrading and elimination time limit (generally no more than 3 years) and the annual transformation and elimination plan, eliminate the projects that cannot be transformed on schedule, and resolutely curb the unreasonable use of energy for high energy consuming projects. We believe that the requirements of relevant policies on benchmark level and time limit show that high energy consuming industries have entered the moment of supply side reform based on efficiency, and the digitization process of high energy consuming industries (steel, petrochemical, chemical, cement, etc.) has accelerated.
policy: the policy is a strong catalyst, and the penetration rate needs to be increased rapidly. in the "14th five year plan" for intelligent manufacturing development, the Ministry of industry and information technology proposes that by 2025, the technical level and market competitiveness of industrial software will be significantly improved, and the market satisfaction rate will exceed 50%. Cultivate more than 150 intelligent manufacturing system solution suppliers with high professional level and strong service ability. we believe that both high-level setting and relevant development planning objectives show that the penetration of industrial software needs to be rapidly improved, and the policy catalyzes the high prosperity of the industry.
[risk tips]
The macro-economy is depressed, major changes have taken place in sector policies, and major changes have taken place in the international environment.
