One of the main expenses and environmental impacts of the modern digital economy stems from the need to maintain cool enough temperatures to operate state of the art data centers. Many data centers utilize the best available technology to reduce their environment impact and lower the costs of cooling equipment with traditional machinery like air conditioning. This has lead to innovations as simple as opening exterior doors and windows, to moves by data giants like Facebook and Google to relocate data centers to very cold climates. However Microsoft is now seeking to revolutionize data cooling by going deep under water with their hardware.
As Microsoft said recently: “50% of us live near the coast. Why doesn’t our data?”
Building massive data centers underwater might sound crazy, but it is exactly something Microsoft is testing with its first submarine data center, dubbed Leona Philpot (a name taken from the company’s popular Halo videogame universe).
The first submerged data center was tested last August, in an enormous steel capsule sunk 30 feet underwater in the Pacific Ocean, about a kilometer off the California coastline. The 8’ wide capsule houses one datacenter computing rack and the exterior of the capsule was enveloped by sensors designed to monitor the underwater environment including pressure, humidity and sea current motion.
105 days later the capsule was recovered and engineers have claimed the experiment was even more successful than they anticipated. According to Microsoft, placing the data center underwater completely eliminated the need for artificial cooling and cut energy costs significantly. Microsoft also explained that while more than half of the world’s population lives within 200 kilometers of the coast, many data centers are inexplicably landlocked far away, creating less than optimal land use policy and extensive data transmission latency which would all be reduced if the data machinery were literally kept offshore.
The goals here are admirable and the early success signals a real possibility for eventual evolution of the way data is stored and accessed. However, it is imperative to keep in mind that these initiatives involve billion dollar companies making high profile bets on distant future technologies. Many factors must be examined more carefully and the diversity of topography, climate and environments from one coastline to another (or even the same coastline miles apart) make it unlikely there will be wide scale adoption of offshore underwater data in the next few years. Still, the company should be applauded for taking the initiative and working toward the sustainable data rich world we all wish to live in for many generations to come.