This isn’t entirely speculative. Microsoft ran Project Natick in 2018, dunking a data center pod in the Scottish North Sea for two years. The servers worked fine. Cooler ocean temperatures kept them running without the cooling towers that land-based facilities require. The startup’s twist is targeting specifically the unused vertical space under wind farms, creating a two-for-one deal: renewable energy above, computing below.
The timing makes sense. Data centers eat enormous amounts of electricity—global capacity is expected to double in the next several years, and finding land for new facilities is getting harder. Communities push back against massive data center campuses. The offshore approach sidesteps those conflicts by moving operations into the ocean, where space constraints are looser and neighbors don’t file complaints about noise or visual impact.
But the ocean is hostile. Saltwater corrodes everything. Currents and wave motion stress hardware. Maintaining underwater servers means sending divers or building remote manipulation systems capable of working in difficult conditions. The startup proposes reinforced steel containers designed to last twenty years, with battery backup for when wind generation dips and specialized cables that keep water out while transmitting power and data. Cooling would be mostly passive—seawater does the work—but engineering the heat exchange systems is still being figured out.
Early reactions have been mixed. Energy consulting firms see a potential market: hundreds of new data centers will be needed over the coming decade, and any strategy that reduces land and energy costs has appeal. The offshore wind industry, which has driven down costs significantly in recent years, might welcome data center hosting as an extra revenue stream from lease areas they already control.
Environmental groups are divided. Direct access to renewable energy cuts carbon emissions. But dumping steel containers and running cables through marine ecosystems isn’t benign. The startup says it will prioritize deeper waters where marine life is less concentrated and conduct impact assessments for each site. Some marine biologists note that underwater structures can function as artificial reefs, actually creating habitat. Whether that offsets the disruption is debatable.
The startup aims to test prototypes within two years, targeting Northern Europe first—regulatory frameworks there are more developed and offshore wind infrastructure is mature. If that goes well, U.S. deployments along the East Coast could follow as that industry expands. Cost parity with land-based facilities might arrive in five to seven years, assuming manufacturing scales up and technology improves.
Government involvement will matter. Maritime regulations weren’t written with underwater data centers in mind. Federal agencies have started preliminary conversations about permitting, but formal rules are far off. There’s also potential national security interest—domestic computing capacity reduces reliance on foreign infrastructure for critical services.
Whether this becomes mainstream or stays experimental depends on those early pilot projects. The logic is appealing: use space nobody wants, draw power from something already there, cut cooling costs dramatically. But marine engineering is unforgiving, and regulatory paths are unclear. The next two years will show whether this is a genuine infrastructure solution or a clever idea that couldn’t survive contact with the ocean.
Leave a comment