① The conference emphasized that central enterprises should plan the development of artificial intelligence in overall work, further promote industrial renewal, and accelerate the layout and development of the intelligent industry. ② A number of central enterprises have already begun to take action early on and have begun racing to gain ground in the AI ecosystem, such as the three major operators.
“Science and Technology Innovation Board Daily”, Feb. 21 — According to the website of the State Assets Administration Commission, on February 19, the State Council's State-owned Assets Administration Commission held a special promotion conference on “AI Empowering Industry Renewal” on artificial intelligence for central enterprises.
The conference believes that speeding up the development of artificial intelligence is an inevitable requirement for state-owned central enterprises to perform their functional mission, seize strategic opportunities, cultivate new quality productivity, and promote high-quality development. Central enterprises should actively embrace the profound changes brought about by artificial intelligence, place accelerating the development of next-generation artificial intelligence in a more prominent position, continuously strengthen innovation strategies, application demonstrations and talent gathering, focus on building artificial intelligence industry clusters, give full play to the advantages of large demand, complete industrial support, and multiple application scenarios, take the lead in seizing artificial intelligence to empower traditional industries, and accelerate the construction of an intelligent economic form driven by data driven, human-robot collaboration, cross-border integration, and co-creation and sharing.
The conference emphasized that central enterprises should plan the development of artificial intelligence in overall work, further promote industrial renewal, and speed up the layout and development of the intelligent industry. It is necessary to consolidate the foundation for development, concentrate the main resources on the areas where they are most needed and have the greatest advantage, speed up the construction of a number of intelligent computing power centers, further deepen open cooperation, and better play the role of a collaborative innovation platform across central enterprises. Launch AI+ special actions to strengthen demand traction, accelerate the empowerment of key industries, build a batch of multi-modal high-quality industrial data sets, and create large models from infrastructure, algorithm tools, intelligent platforms to solutions to empower the industrial ecosystem.
At the meeting, representatives of leading companies and central enterprises concerned with artificial intelligence made exchange statements. Ten central enterprises signed a proposal stating that they will take the initiative to open up artificial intelligence application scenarios to the public.
At the end of January, Secretary General of the State Council's State-owned Assets Administration Commission and press spokesman Zhuang Shuxin explained in response to an inquiry that since last year, state-owned central enterprises have taken the development of strategic emerging industries and future industries as a leading and global task, focusing on strategic emerging industries such as next-generation mobile communications, artificial intelligence, biotechnology, new materials, etc., and future industries such as future information to launch strategic emerging industries and future industry launch actions.
The next step will be to plan and promote a number of major projects, select “100 projects” to fight new industries, build a number of strategic emerging industry clusters, and implement special actions such as artificial intelligence AI+.
▌The three major operators took the lead, and central enterprises started “AI” in the horse racing industry
In fact, a number of central enterprises have already taken action early on, and have begun racing to gain ground in the AI ecosystem, such as the three major operators.
Among them, the “Science and Technology Innovation Board Daily” reporter learned today from people familiar with the matter that China Telecom is planning to build an intelligent computing center in Shanghai that can support the training of models with large trillion parameters. A single pool has built 10,000 cards of domestic computing power. It is the first domestic hyperscale computing power liquid cooling cluster that supports 10,000 cards in a single pool.
Hu Yubo, deputy general manager of China Mobile's information disclosure department, said earlier that China Mobile has deployed corresponding intelligent computing servers in recent years and is planning to build the largest intelligent computing center in Asia. In terms of data, the daily collection volume exceeds 5 PB, and the core data scale exceeds 650 PB.
China Unicom, on the other hand, is committed to building a computing power boutique network with rich computing power, abundant capacity, multi-cloud collaboration, and integrated computing networks. Its future plans mainly include three aspects: (1) Enriching the supply of multi-level computing power and improving the multi-level architecture. (2) Improve the computing power transmission network capacity, build a strong capacity, intelligent and open all-optical transmission base; optimize the network architecture layout, and build a low latency plane for East and West computing. (3) Optimize the network scheduling system.
In addition, the three major operators, China Mobile, China Telecom, and China Unicom, jointly released the “White Paper on Liquid Cooling Technology for Telecom Operators” in June 2023. Among them, a three-year development vision is proposed. The three major operators will carry out technical verification in 2023; carry out large-scale tests in 2024 to pilot liquid cooling technology on a 10% scale; in 2025, large-scale applications will be carried out, and more than 50% of projects will apply liquid cooling technology to jointly promote the formation of a high-quality development pattern with unified standards, perfect ecology, optimal cost, and large-scale applications.
As the demand for artificial intelligence resources increases day by day, the continued shortage of computing power networks will further drive the demand for hardware infrastructure construction. In addition to operators, the artificial intelligence industry chain also includes many important companies such as optical modules, switches, and chips.
According to incomplete statistics from the “Science and Technology Innovation Board Daily”, central enterprises related to artificial intelligence, especially infrastructure, in A-shares include:
