Related to AI computing power! This new infrastructure is included in the government work report for the first time (with stock suggestions)

“Computing and Power Collaboration” is written into the government work report for the first time.

On March 5th, the “Government Work Report” proposed to “implement ultra-large-scale intelligent computing clusters, computing and power collaboration, and other new infrastructure projects, strengthen nationwide integrated computing power monitoring and scheduling, and support the development of public clouds.”

“Computing and power collaboration” was included in the government work report for the first time, explicitly listed as a new infrastructure project, marking its transition from local pilot projects and departmental policies to a formal national strategic deployment.

Chen Changsheng, a member of the drafting team of the “Government Work Report” and Deputy Director of the General Office of the State Council, stated at a briefing by the State Council Information Office on March 5th, “To lay a solid foundation for AI development, we must first strengthen the infrastructure foundation. There is a saying online that AI’s end is energy. We need to leverage the advantages of the national power grid system and further implement the construction of ultra-large-scale intelligent computing clusters and computing-power collaboration as new infrastructure. From hardware and technical perspectives, we should further support enhancing large model capabilities and computing power, and deploy multiple pathways and routes for embodied intelligence, world models, and more.”

“Computing and power collaboration” refers to the organic integration of computing infrastructure and power systems through digital technology, intelligent algorithms, and communication networks. It aims to achieve dynamic matching and optimal allocation of computing and electrical resources through intelligent scheduling, enabling collaborative interaction in resource dispatch, operational management, and demand response, thereby improving energy utilization efficiency, ensuring stable computing supply, and promoting low-carbon development.

Behind computing power is electricity

Computing power is the core engine of the digital economy, but data centers (intelligent computing centers) are high-energy-consuming loads. According to China Academy of Information and Communications Technology (CAICT), from 2019 to 2024, China’s data center electricity consumption increased from 82.4 billion kWh to 166 billion kWh, with an average annual compound growth rate of 15%.

Additionally, overseas expansion of computing power is becoming a new industry trend. According to OpenRouter, the world’s largest AI model API aggregation platform, from February 16th to 22nd, the top ten models consumed about 8.7 trillion tokens, with Chinese models accounting for 5.3 trillion tokens, or 61%. The overseas expansion of computing power essentially reflects the export of electricity.

IDC forecasts that from 2025 to 2028, China’s intelligent computing power will grow from 1037.3 EFLOPS to 2781.9 EFLOPS, with an average annual growth rate of 38.9%. Assuming annual growth rates of 25%, 35%, and 45% respectively, by 2030, China’s intelligent computing power will reach 3166 EFLOPS, 4651 EFLOPS, and 6649 EFLOPS, with data center electricity consumption reaching 390.7 billion kWh, 574.1 billion kWh, and 820.6 billion kWh.

Computing and power collaboration ensures energy supply

Faced with the high energy consumption of data centers, as early as December 2023, the “Implementation Opinions on Deepening the ‘East Data West Computing’ Project and Accelerating the Construction of a National Integrated Computing Power Network” first mentioned “computing and power collaboration.” In July 2024, the “Green and Low-Carbon Development Special Action Plan for Data Centers” and the “Action Plan for Accelerating the Construction of a New Power System (2024–2027)” proposed a series of specific measures to promote computing and power collaboration.

Last May, the National Development and Reform Commission and the National Energy Administration jointly issued the “Notice on Steadily Promoting Green Power Direct Connection Development.” The latest data from the National Energy Administration shows that 84 green power direct connection projects have been approved nationwide, with a total new energy installed capacity of 32.59 GW, covering projects such as data center power supply downstream.

Guosheng Securities states that the core of computing and power collaboration lies in guiding high-quality new electricity consumption, such as “green power direct connection” and “smart microgrids,” to more precisely match the production of fluctuating renewable energy with computing load in space and time, thereby ensuring energy supply while accelerating green energy transformation.

Exploring new models of computing and power collaboration is not only an important practice for implementing the “dual carbon” goals and the coordinated promotion of the Digital China strategy but also an innovative path to solve the high energy consumption problem in the computing industry.

Companies involved in computing and power collaboration

On the listed company level, based on incomplete public information compiled by Securities Times·Data Treasure, including power operators, data center service providers, and electrical equipment companies, more than 20 companies have clarified their computing and power collaboration businesses.

Kehua Data stated in its 2025 semi-annual report that computing and power collaboration is an inevitable trend. Its strategic layout of “Intelligent Computing Center” and “New Energy” has demonstrated its collaborative advantages, and the integration of “Intelligent Computing + Photovoltaic Storage” has become a core competitive edge.

Tongli Tianqi mentioned during research activities that the company is closely following the trend of computing and power collaboration. In March 2025, it signed a strategic agreement with Qingyang City government to build a 2 GWh energy storage equipment production line and a 1 GWh lithium iron phosphate energy storage power station, exploring intelligent scheduling models for computing and power.

Huazi Technology stated on the investor interaction platform that, relying on core advantages such as integrated source-grid-load-storage, computing and power collaboration, and zero-carbon power supply, the company is actively expanding related green energy markets. By 2026, it will focus on national computing hub nodes, increase efforts in key regional expansion and project reserves, and promote rapid business deployment.

Zhongheng Electric told investors that the company is continuously increasing R&D efforts in data center power supply, aiming to develop ultra-efficient, highly reliable, and intelligent third-generation HVDC products to meet the power needs of high-density servers in intelligent computing centers, helping achieve the “Three Zero” goals (zero failures, zero engineering, zero losses) and ultimately realize computing and power collaboration.

China Energy Construction Corporation stated that by building new supporting projects for renewable energy, and adopting a “green power direct supply + smart grid + computing and power collaboration” integrated model, it provides long-term, stable, and low-cost green electricity for data centers.

Several power operators are also directly involved in data center fields, such as Jinkai New Energy investing in the Xinjiang Changji 5000P intelligent computing center and signing contracts for the Ulanqab intelligent computing center project, and Yunnan Energy Holding planning to participate in the equity of Chuangtian Computing and acquire Zhengzhou Heying Data Holdings, among others.

(Source: Securities Times Network)