Google has created artificial intelligence that's able to save the amount of electricity it uses to power its data centres.
Using machine learning developed by the firm's AI research company, DeepMind, it was possible to reduce the energy used for cooling the centres by a staggering 40 per cent.
By applying machine learning to its own centres, which power Google Search, Gmail, YouTube and all of Google's services, it was able to improve their efficiency.
The algorithms and methods used could also be transferred to air conditioning systems in large manufacturing plants or, on an even larger scale, to reduce wastage in the energy grid.
"What we've been trying to do is build a better predictive model that essentially uses less energy to power the cooling system by more accurately predicting when the incoming compute load is likely to land," Mustafa Suleyman, the co-founder of DeepMind told WIRED.
"Also in real time we're adjusting the parameters of the cooling system so that it more closely matches the demand from the compute processing."
In essence the system has been created to respond to the demand that is being put on it and reduce the amount of electricity needed when it is possible to do so.
Suleyman's Dougal team – a division of DeepMind building projects for direct use within Google – created the algorithms using deep neural networks. The network type aims to mimic the functionalities of the brain and have been used in everything from creating an artificial Donald Trump to treating serious diseases such as Alzheimer's.
The DeepMind team collected five years worth of data collected by data centres and created a prediction model for how much energy would be needed by the data centre based on the amount of server usage that was likely. Each neural network was fed data on temperatures, power usage, pump speeds and more.
By using the large data sets, the machine learning was able to be "trained" and retain more examples of how the centres' work than a human would be able to.
"Conventionally a human manually tweaks a lot of the knobs that control the operation of the data centre," Suleyman explained. "There's obviously a lot of variation in performance across all the data centres because each human performs quite different."
When controlling a running data centre, in recent months, Google said the AI was able to "consistently achieve a 40 per cent reduction in the amount of energy used for cooling".
The algorithms created were to be a general learning artificial intelligence. This means it may be possible to apply it to other scenarios. "There's lots of other applications outside of Google," Suleyman said.
"We think there's lots of potential to apply this to large scale energy distribution, so we're giving it some thought and are in early discussions with a number of people on that."
This article was originally published by WIRED UK
