At the ninth China International Import Expo, multinational companies presented technologies aimed at three demands linked to AI growth in China: more efficient data centers, a more flexible power system and industrial AI that works reliably in production. The exhibits and executives’ comments describe company plans and claims, not independently verified performance results.
Data centers and electricity demand
Schneider Electric showed an AI data-center offering covering power, cooling, deployment and operations. According to staff at its stand, the company’s factory-prefabricated IMDC POD 2026 modular data center could shorten on-site construction time by about 50% and achieve a power usage effectiveness (PUE) as low as 1.09. These are company-stated figures, not measurements reported from an operating facility.
Siemens presented its Data Center Solution 5.0+. Xiao Song, president of Siemens Greater China and chairman of Siemens China, said it responds to the need to coordinate computing and energy use, including a shift toward direct-current power architecture and liquid cooling. He described a longer-term vision in which AI data centers could help balance the grid rather than act only as fixed electricity loads.
Citing an International Energy Agency forecast, the original report said global data-center electricity consumption could reach 950 billion kilowatt-hours by 2030, or 3% of global electricity demand. It put consumption by AI-specific data centers at about 465 billion kilowatt-hours by then. Ge Ju, deputy director of the Department of Development and Planning at China’s National Energy Administration, said at a recent press conference that China’s annual electricity demand could rise by around 600 billion kilowatt-hours on average during the 15th Five-Year Plan period, partly driven by emerging industries such as computing.
Siemens Energy also displayed equipment intended to support grid stability and flexible generation. Its SGT-9000HL gas turbine was among the exhibits; the report described its potential roles in combined-cycle generation, combined heat and power, and responding to fluctuations in renewable output.
Industrial AI beyond demonstrations
Honeywell Technology unveiled a China AI research institute during the expo. Its Forge connected platform, which incorporates AI, had launched in China in July, according to the report. Yu Feng, president of Honeywell Technology Greater China, said the company was moving from displaying individual products toward offering cross-business solutions for Chinese customers.
Shi An, president of Rockwell Automation China, argued that industrial AI needs an understanding of manufacturing processes—not just data and models—to solve production problems. Rockwell showed two new-generation controllers at the expo. His comments and the displays underline a distinction between demonstrating AI tools and establishing that they improve operations across factories.
Original report (The Paper), published November 10, 2026.