The 2026 China Computing Power Conference is being held from September 11 to 13 in Langfang, Hebei. The Beijing-Tianjin-Hebei Forum on AI Computing and Advanced Manufacturing, one of the conference's main events, took place on September 11.

During the forum, the Beijing-Tianjin-Hebei AI computing supply and demand lists were formally published, creating a bridge for precise matching and project cooperation between suppliers and users of computing power, and pushing AI computing and advanced manufacturing in the region closer together.

Supply and Demand Lists Drive Precise Matching

On the supply side, the lists bring together providers including Zhuozhou Hengzhi Yuntong, the Beijing Public Computing Power Platform for Artificial Intelligence, and China Telecom's Beijing-Tianjin-Hebei AI data center. Together they form a supply structure led by large-scale AI computing clusters, supplemented by general-purpose computing power located nearby, with storage capacity built out alongside compute.

On the demand side, the lists gather 64 requests for computing power spanning steel, equipment manufacturing, auto parts, pharmaceuticals, government services, and cultural and creative industries.

Those requests range from large-scale model training and inference to smaller, more elastic needs in areas such as AI-based visual inspection and industrial simulation, reflecting a wide spread of use cases and tiers of demand.

"The key to coordinating computing power is matching supply with demand," said an official with the Hebei Provincial Department of Industry and Information Technology. Industry and IT departments in the three jurisdictions will use the release of the lists as a starting point to set up a routine tracking mechanism for computing power supply and demand, keep organizing resource matching sessions between the two sides, and speed up deals covering compute procurement, compute leasing, and joint construction projects.

Three Jurisdictions Sign "Galaxy Computing Corridor" Agreement

During the forum, the Beijing Municipal Bureau of Economy and Information Technology, the Tianjin Municipal Bureau of Industry and Information Technology, and the Hebei Provincial Department of Industry and Information Technology signed a cooperation agreement on jointly building the Beijing-Tianjin-Hebei "Galaxy Computing Corridor" for coordinated computing power development.

Under the agreement, the three will cooperate on building the corridor itself, linking up computing power scheduling, combining green energy with computing, growing the computing power industry, developing shared standards and technology, and maintaining security baselines, with the aim of creating the northern backbone of the national computing power network.

The three will accelerate construction of the "Galaxy Computing Corridor" project, build a computing power supply corridor ringing Beijing, and extend the industry's upstream and downstream chains.

Looking ahead, the region plans coordinated development across chips, storage, communications, complete server systems, liquid cooling, data centers, computing power scheduling, large models, and application scenarios, building out a computing power industry that covers the full supply chain.

Through regional coordination, the three hope to establish an end-to-end model running from computing power infrastructure construction to resource scheduling and industry deployment, improving overall utilization of the region's computing resources.

Beijing-Tianjin-Hebei AI Computing Capacity Keeps Growing

According to information presented at the forum, the three jurisdictions are seizing on the national East Data West Computing project to accelerate construction of the Beijing-Tianjin-Hebei national hub node in the country's integrated computing power network.

Beijing, Tianjin and Hebei together account for roughly 30% of China's total AI computing capacity, the forum disclosed. Hebei holds particular strengths in raw capacity and in opening up application scenarios.

As of the end of June 2026, five projects in Hebei had been selected as national model cases for new-type data centers, three had been shortlisted in the national computing power foundation-strengthening program, and the province had cultivated 22 national green data centers, among the highest counts in the country.

Over the same period, Hebei's AI computing capacity reached 556.1 EFLOPS and its in-service rack count stood at 2.504 million standard racks, both first nationally.

EFLOPS is a measure of floating-point performance; one EFLOPS equals a quintillion floating-point operations per second. The growth in AI computing capacity and rack numbers gives the region the base it needs to host AI training, inference, and industrial AI applications.

Demand for Computing Power Reaches into Manufacturing

The Beijing-Tianjin-Hebei Forum on AI Computing and Advanced Manufacturing drew more than 200 participants from government agencies, research institutes, and companies across the computing power supply chain.

Companies at the forum said manufacturing is shifting from straightforward digitization to AI-driven upgrades, with demand for computing power expanding beyond centralized large-scale training into specific use cases such as visual inspection, industrial simulation, predictive maintenance, and production process optimization.

Jiang Shaoqiu, product director at Rongke Lianchuang Information Technology Co., said the supply and demand lists give companies a clearer picture of what Hebei's manufacturers need, and that his firm will step up matchmaking efforts and pursue more computing power use cases.

Hu Kun, senior business manager for computing power interconnection at the China Academy of Information and Communications Technology, said computing power interconnection has moved from the network-building stage to the network-using stage. The next step, he said, is to deepen computing coordination across supply chains such as new-energy intelligent connected vehicles and robotics, and to produce a set of "computing power plus manufacturing" flagship applications that can be replicated elsewhere.

He recommended that the region cultivate a specialized computing services ecosystem offering manufacturers compute on demand, and that it use computing power node platforms to deliver inclusive computing power, helping small and mid-sized tech companies obtain AI computing at lower cost.

Computing power is a key productive force in the digital economy and the infrastructure underpinning AI development. As the region's mechanisms for matching computing supply and demand mature and the "Galaxy Computing Corridor" project advances, its computing resources are expected to move toward unified scheduling and more efficient coordination, drawing computing infrastructure and advanced manufacturing closer together.