Microsoft has recently experimented with underwater data centers. They began by dropping a server rack with the power of roughly 300 PCs into the ocean off of the coast of California. This experiment was carried out due to the fact that current data centers have proved to be inefficient. According to Microsoft, there could also be a number of environmental benefits. The goal of the company is to be able to utilize these data centers with a product of zero emissions.
I believe that these underwater data centers could prove to be extremely effective. First, there is great value in improving the overall efficiency of these centers. The fact that these pods would be placed in areas that are closer to humans than those above land would greatly improve the latency of current computers. Additionally, current data centers utilize a lot of energy in an attempt to cool off their systems. However, underwater, the ocean water would naturally cool these centers. This would create a net emission of zero, and greatly help the environment.
Some could argue that the heat given off by these centers could harm underwater wildlife. However, Microsoft has stated that these centers have a net heat emission of zero, which if true, would not affect wildlife. Additionally, these data centers are built from recycled materials. Microsoft believes this could greatly improve the carbon footprint compared to above water centers. The pros seem to greatly outweigh the cons in this experiment. This first experiment is a small step in an even larger plan for Microsoft. "Our first experiment was like dipping our pinkie toe in the water, and now we're going for the big toe," said Peter Lee, corporate vice president of Microsoft Research NExT. The company plans to move on another project soon, in which they hope to increase the effectiveness of the data centers. They will equip the next batch of pods with turbines, which will allow the data centers to utilize the wave energy in order to operate.
Microsoft hopes to be able to create fully operating underwater data centers in roughly 90 days. Current data centers usually take up to two years to build. In my opinion, this discovery could lead to something we take for granted ten years from now. The efficiency of these under water centers could exponentially improve data center speed, and decrease harmful effects on the environment. Although these centers are not imminently going to be put to full use, I look forward to seeing if one day Microsoft can find a way to make them work on a large scale.