China's energy storage industry is pivoting toward artificial intelligence data centers (AIDCs), with "computing power plus storage" emerging as a new growth driver as AI electricity demand surges.

The China Energy Storage Alliance (CNESA) reported that Chinese companies signed overseas energy storage orders totaling 298 GWh in the first half of 2026, up 83% year on year, and identified computing-power-plus-storage as a new growth point for the sector.

Industry forecasts project lithium battery shipments for AIDCs will grow from 15 GWh in 2025 to 69 GWh in 2027, and potentially exceed 300 GWh by 2030. China's AIDC electricity consumption is expected to post a compound annual growth rate above 40% from 2026 to 2030, while global computing center storage demand is projected to reach about 200 GW by 2030. By comparison, global cumulative installed energy storage capacity stood at approximately 270 GW as of the end of 2025, suggesting computing-power-plus-storage could create incremental space nearly equivalent to rebuilding the entire global storage market.

Industry consensus holds that computing centers will be the fastest-growing electricity consumption segment over the next five years, accounting for more than 8% of total social electricity use by the end of the 15th "Five-Year Plan" period. Electricity costs represent a relatively limited share of total investment for computing centers, giving them far higher tolerance for electricity price increases than ordinary users. This makes integrated computing-power scenarios capable of delivering more substantial profit potential for storage stations than standalone storage.

Through peak-valley arbitrage, demand control and demand response models, paired storage systems can help AIDCs reduce comprehensive electricity costs by more than 30%, transforming storage from a cost center into a profit center.

At the April 2026 Energy Storage International Summit, nearly all leading storage companies reserved their core exhibition space for AIDC storage products. Companies from CATL to Trina Solar showcased full-industry-chain AIDC offerings spanning battery cells, modules, storage systems, power distribution equipment and energy management software.

First-half 2026 earnings reports from listed storage companies showed Shuangdeng Group's AIDC lithium battery revenue grew more than 400-fold year on year, while Sunwoda's AIDC storage orders rose nearly 30-fold. Sungrow Power Supply reported approximately 2 GWh in AIDC storage orders on hand and stated it aims to secure data center storage orders in 2026.

At the project level, Shuangdeng Group won a bid for China's first large-scale AIDC park integrated storage project featuring direct green power connection and source-grid-load-storage coordination, with a capacity of 120 MW/240 MWh. Haibosichuang announced plans to build a lithium-sodium hybrid plus storage-computing integration project in Baotou, Inner Mongolia autonomous region, exploring a joint operation model combining energy storage with intelligent computing centers.

On the policy front, China's 2026 Government Work Report included "computing-power coordination" for the first time, listing it alongside "ultra-large-scale intelligent computing clusters" as key new infrastructure projects. A joint action plan issued by the National Development and Reform Commission and three other government bodies explicitly encourages computing facilities to deploy grid-forming storage to enhance power supply stability and active support for the power system, and outlines specific technical directions addressing the high-power, high-fluctuation load characteristics of intelligent computing centers. The National Energy Administration said in response to a National People's Congress proposal that future new-type energy storage pilot programs will prioritize AIDC storage application scenarios and that it will organize research on AIDC storage standards.

The significance of computing-power-plus-storage for the storage industry may extend beyond opening a new demand window. Traditional standalone storage revenue models rely heavily on peak-valley price spreads and capacity leasing, with relatively narrow income sources that are highly susceptible to policy adjustments. Computing-power-plus-storage embeds storage into the computing production process, so storage value is no longer limited to power system regulation but directly participates in the computing cost structure. Through direct green power connection and storage peak-shaving, data centers can achieve lower comprehensive energy costs and higher green power consumption ratios. The principle of "computing follows power, power adjusts with computing" means storage companies will need to understand both electricity markets and computing production logic, placing entirely new demands on the industry's capability structure.

However, a gap remains between a new growth point and a stable growth engine. Misalignment between computing and power planning, missing market mechanisms and an incomplete industry standards system remain the main bottlenecks to large-scale commercialization. Storage operations and maintenance at computing parks also face unique challenges: long-term idle standing and full-charge float charging of lithium batteries accelerate cell consistency degradation, while computing parks generally lack dedicated storage O&M teams. Whether these issues can be effectively resolved will determine how much of the computing-power-plus-storage market potential can be truly realized.