Shanghai Launches First Commercial underwater Data Center
Shanghai, China – A Chinese company is pioneering a novel approach to data center cooling and energy efficiency by deploying the first commercial underwater data center off the coast of Shanghai. The project, developed by Highlander with state construction companies, aims to drastically reduce the energy demands of computing technology by leveraging the natural cooling properties of the ocean.
Data centers,essential for powering websites and applications – and increasingly,artificial intelligence – are notoriously energy-intensive. This new facility offers a potential solution by utilizing ocean currents to keep servers cool, a method far more efficient than conventional air or water cooling systems used in land-based centers.
“Underwater operation has the benefits of…” stated Yang Ye, a highlander spokesman.
The submerged module is projected to save approximately 90 percent of the energy typically consumed for cooling. it will initially serve clients including China Telecom and a state-backed AI computing company, aligning with the government’s broader efforts to minimize the carbon footprint of its data centers.
The data center consists of prefabricated modules assembled onshore and then installed at sea, powered largely by nearby offshore wind farms. Servers are shielded from corrosion and maintenance is facilitated by an elevator system extending above the waterline.
While inspired by earlier tests conducted by Microsoft off the coast of scotland in 2018, the Shanghai project represents one of the first commercial deployments of this technology globally.
However, the long-term ecological impacts remain under investigation. Experts note that the heat discharged from the underwater facility could attract some marine species while deterring others, and further research is needed to fully understand these effects. It is anticipated that underwater data centers will likely complement, rather than replace, traditional facilities, possibly serving specialized computing needs.