Can submerged data centers power ai's growth sustainably?

The rapid expansion of artificial intelligence is increasing demand for data centers, prompting companies to seek alternatives to traditional land-based facilities due to concerns over energy use, water consumption, and carbon emissions. Some developers are now considering the ocean as a potential site for AI infrastructure, believing that underwater data centers could offer improved energy efficiency and cooling while reducing reliance on fresh water and land. However, these centers do not resolve underlying environmental issues such as energy consumption and carbon emissions, and they raise new concerns about potential harm to marine ecosystems and regulatory challenges. Microsoft, which launched a research project in 2015 to explore underwater data centers, deployed a sealed facility on the seafloor near Scotland’s Orkney Islands in 2018. The center, containing 864 servers, was connected to shore via an underwater cable. After two years, Microsoft reported that the servers in the underwater facility failed at about one-eighth the rate of comparable land-based servers, though the company has not disclosed the exact reasons for this improvement. The company ended the project in 2024 without announcing the reasons, but analysts suggest regulatory hurdles, such as the need for environmental permits, and the desire for faster equipment upgrades may have influenced the decision. Microsoft’s withdrawal highlights the complex challenges facing underwater data centers, including technical, economic, and environmental factors that must be addressed before they can be considered a sustainable alternative to traditional data centers.