Pranav Hotkar
Beijing, China - June 9, 2026 - China has launched what is being described as the world’s first operational underwater data center powered directly by offshore wind energy, highlighting growing global efforts to develop alternative infrastructure models for energy-intensive AI and cloud computing workloads.
The project, located near Shanghai, combines submerged data center modules with a nearby offshore wind farm to create a low-carbon digital infrastructure platform designed to reduce cooling requirements and improve energy efficiency for high-density computing environments.
According to reports, the underwater facility has now begun commercial operations following testing and deployment phases. Operators say the system uses surrounding seawater for natural cooling, significantly reducing the need for traditional mechanical cooling systems that account for a major share of data center electricity consumption globally.
The development comes as AI infrastructure growth sharply increases energy demand worldwide. Large GPU clusters supporting generative AI training and inference are driving unprecedented electricity consumption, prompting operators and governments to explore new approaches to power delivery, cooling, and infrastructure efficiency.
China has rapidly expanded its AI and digital infrastructure ambitions over the past several years, with government-backed initiatives supporting hyperscale cloud development, advanced semiconductor manufacturing, and renewable-powered data center projects.
The underwater data center initiative forms part of a broader trend toward unconventional infrastructure strategies aimed at reducing operational costs and environmental impact. Similar underwater data center concepts have previously been explored by companies, including Microsoft through its Project Natick research initiative, though China’s deployment is believed to be the first directly integrated with offshore wind power at a commercial scale.
Industry analysts say underwater facilities could offer several advantages for future AI infrastructure deployments, including lower cooling costs, improved energy efficiency, reduced land usage, and proximity to coastal fiber connectivity routes.
The project also reflects increasing convergence between renewable energy infrastructure and hyperscale computing development. Access to large-scale renewable electricity has become one of the most critical factors shaping where next-generation AI campuses are being built globally.
China remains one of the world’s largest offshore wind markets, and pairing renewable generation directly with data center infrastructure is increasingly viewed as a potential strategy for supporting future gigawatt-scale AI deployments while managing grid pressure and emissions.
While technical details regarding total IT capacity and GPU deployment levels have not been fully disclosed, reports indicate the underwater facility is designed to support AI processing, cloud services, and high-performance computing applications.
The launch further underscores how data center operators worldwide are experimenting with new cooling architectures and energy systems as traditional infrastructure models face mounting pressure from rising AI compute density and electricity consumption.
As global AI infrastructure investment accelerates, alternative concepts such as underwater facilities, liquid cooling, modular nuclear integration, and renewable-powered campuses are increasingly moving from experimental research into commercial deployment.
