Machine learning optimizes Google data centers' PUE

Did you know that Google has been calculating its data centers’ PUE data every five minutes for over five years, including 19 different variables such as cooling tower speed, processing water temperature, pump speed, outside air temperature, humidity, etc? The Internet search giant has by now reportedly compiled "hundreds of millions" of data points to analyze and put to use in making its data center energy operations better. GigaOM's Katie Fehrenbacher has the story of how "neural networks are the latest way Google is slashing energy consumption from its data centers" and how "only a [sophisticated] computer model could crunch all the data Google has collected on its data center energy consumption."

Full story:Google is harnessing machine learning to cut data center energy(gigaom.com)

See also: Google's data center building boom thunders on

More: Fiber micro-trenching at Google