China’s Computing-Power-Electricity Synergy Push
In a quiet corner of Gui’an New Area in Guizhou Province, a 40-foot shipping container sits unassuming on open ground. Inside, there are no roaring fans or tangled cables—just servers quietly humming, submerged in transparent cooling fluid. This is Guizhou’s first “computing-power container” (算力方舱), an immersion liquid-cooled data center built by Southern Power Grid’s Guizhou subsidiary and put into operation in early August 2026. It represents the tangible realization of China’s ambitious computing-power and electricity synergy (算电协同) strategy, a policy initiative first elevated to national significance when it was written into the 2026 Government Work Report during the National People’s Congress in March.
The compact modular data center was showcased at the 2026 China International Big Data Industry Expo held in Guiyang from August 28-30, where Southern Power Grid hosted a themed exchange event on “Electricity-Computing Symbiosis, Digital Future.” The initiative addresses a fundamental challenge of the AI era: the enormous electricity consumption required to power the nation’s rapidly expanding computing infrastructure.
From Policy Blueprint to Industrial Reality
The concept of computing-power and electricity synergy evolved through several stages before reaching national strategic status. It was first mentioned in December 2023 in a State Council implementation opinion on the “East-Data-West-Computing” (东数西算) program, followed by specific development measures in July 2024. In May 2025, authorities issued a notice on orderly promoting green electricity direct connections. The decisive moment came in March 2026, when “算电协同” was written into the Government Work Report for the first time, explicitly listed as a new infrastructure project alongside ultra-large-scale intelligent computing clusters.
As Xinhua News reported, the Government Work Report called for implementing new infrastructure projects such as ultra-large-scale intelligent computing clusters and computing-power-electricity synergy, strengthening national integrated computing power monitoring and dispatch, and supporting public cloud development. This marked the first time “new infrastructure project” appeared as a new term in the report.
Chen Changsheng, deputy director of the State Council Research Office and a member of the Government Work Report drafting group, framed the challenge directly: “There’s a saying online that the end of AI is energy. We need to leverage the advantages of the national power grid system and further implement new infrastructure construction such as ultra-large-scale intelligent computing clusters and computing-power-electricity synergy.”
The Liquid-Cooled Computing Container
The centerpiece of the current phase is the immersion liquid-cooled data center deployed at the Southern Energy Big Data Center in Gui’an New Area. As China News Service documented, the container integrates four immersion liquid-cooled racks totaling 80kW of IT power consumption, achieving a power usage effectiveness (PUE) of ≤1.15—a significant improvement over traditional air-cooled data centers. Construction time was reduced to just three months, with initial investment savings of 40-50 percent.
The container uses specialized liquids such as fluorinated fluids and synthetic oils to directly envelop servers and carry away heat, with external closed cooling towers completing the heat circulation. The factory-prefabricated design allows for plug-and-play deployment at substations and zero-carbon parks, as C114 Communications Network detailed in its coverage of the 2026 Big Data Expo.
Wang Pengyu, deputy general manager of the resource operations team at Guizhou Electric Power Company’s Digital Intelligence Center, described the container as a “computing accelerator” for grid operations. Massive images from substation inspections and drone patrols of transmission lines that previously needed cloud uploads—potentially causing lag or queueing—can now be handled locally through edge processing, greatly improving response speed.
Breaking Down Barriers: Electricity-Carbon-Computing Synergy
Beyond the hardware innovation, China is pioneering institutional mechanisms to coordinate electricity, computing power, and carbon emissions within a unified market system. In May 2026, the first batch of electricity-carbon-computing synergy matching transactions (电碳算协同撮合交易) began pilot operations in Guizhou, placing all three elements in the same market framework for coordinated allocation.
Users can select green computing resources during peak wind and solar generation periods—similar to ordering a ride-hailing service—reducing computing costs while improving renewable energy consumption. Through the electricity-carbon-computing synergy operation system, some computing users have already seen comprehensive computing cost reductions of over 10 percent.
Xuan Liang, deputy general manager of Southern Power Grid’s Digitalization Department, explained the company’s four-dimensional coordination approach: “Relying on the abundant hydro, wind, and solar clean energy resources of the five southern provinces, Southern Power Grid is seizing the opportunity of national integrated computing network construction, promoting four-dimensional coordination of planning, construction, dispatch, and trading, breaking down barriers between electricity, computing power, and carbon emission factors.”
Hu Rong, general manager of Southern Power Grid’s Digital Grid Research Institute, articulated the vision of a closed-loop ecosystem: “Electricity-carbon-computing synergy is not a one-way ‘power supply to computing centers,’ but an organic ecosystem where energy flows and data flows drive each other. AI both safeguards the large power grid and dispatches computing resources; computing power consumes green electricity to generate tokens, tokens generate AI capabilities based on data, which then serve the energy system.”
National Expansion and Market Impact
The practice is expanding beyond Guizhou. Inner Mongolia and Ningxia, as “East-Data-West-Computing” hub nodes, are advancing direct connections between wind and solar bases and intelligent computing centers, with green electricity dedicated lines reaching server rooms. Multiple regions are conducting computing load peak-shaving tests, shifting non-real-time rendering and model pre-training tasks to periods of abundant renewable energy output.
The economic stakes are substantial. Financial Street Securities Chief Economist Zhang Yi estimates the synergy model could create at least a trillion-yuan market increment by driving development across energy storage, smart grids, data centers, and green electricity operations. As China Energy News reported, Southern Power Grid’s cross-regional computing migration has already achieved a 16 percent reduction in total electricity costs for Guangzhou AI training tasks migrated to Gui’an, with comprehensive computing electricity costs reduced by 16-33 percent through coordinated scheduling.
Challenges Ahead
Despite the momentum, experts caution that computing-power-electricity synergy remains in its early stages. The spatial mismatch between computing demand and energy resources is a persistent challenge—eastern regions account for over 60 percent of computing demand but less than 20 percent of energy resources. As China Energy News reported in its Two Sessions analysis, national legislators and political advisors have called for establishing unified evaluation indicators, clarifying the market value of flexible computing loads, improving revenue sharing mechanisms, and encouraging multi-party participation from data centers, power grids, and computing users.
Chinese Academy of Engineering Academician Wang Jian highlighted the shared characteristic of electricity and computing power: peak and valley periods. “If these peaks and valleys can be well coordinated, the energy needed for computing will be greatly optimized,” he noted.
The Road to 2030
The scale of the challenge is significant. China’s data center annual electricity consumption grew from 82.4 billion kWh in 2019 to 166 billion kWh in 2024, according to China Academy of Information and Communications Technology data. By 2030, China’s computing power electricity consumption is projected to reach 480 billion kWh, with carbon emissions exceeding 200 million tons.
China brings unique advantages to this challenge: the world’s largest power grid system, wind and solar installed capacity of 1.84 billion kW accounting for nearly half of global installations, and computing scale ranking second globally. As the China Power industry portal noted, nearly 4 out of every 10 kWh consumed nationwide is green electricity.
The 2026 Big Data Expo showcased how these elements are converging. Southern Power Grid’s “Electricity-Computing Symbiosis, Digital Future” event at the expo, as CNR reported, focused on releasing new achievements of the electricity-carbon-computing synergy operation system and launching the energy industry trusted data space pilot.
As the initiative continues to evolve from policy blueprint to industrial reality, the synergy between computing power and electricity is poised to become a defining feature of China’s digital economy—one that could determine the nation’s competitive position in the global AI race. The question now is how quickly the remaining institutional and technical barriers can be overcome to unlock the full potential of this trillion-yuan opportunity.