• Fayaz Ahmed

How DeepMind’s AI algorithm reduces Google Data Centre Energy Bill




I am sure you must have noticed your laptop getting hot while using it for longer periods, similarly when thousands of servers at Google Data Centre that power popular services including Google Search, Gmail and YouTube- also generate a lot of heat that must be removed to keep the servers running. This cooling is typically accomplished via large industrial equipment such as pumps, chillers and cooling towers consuming significant amounts of energy. Therefore, even minor improvements in the energy efficiency of these data centers would provide significant energy savings and reduce CO2 emissions to help combat climate change.



Thanks to Machine learning technology that we are able to seamlessly benefit from smartphone assistants, image recognition and translation in our everyday lives. Despite lots of success stories of machine learning it wasn’t sure that machine learning can also help us to tackle some of the world’s most challenging physical problems - such as energy consumption until a British artificial intelligence (AI) company DeepMind, acquired by Google in 2014, developed an AI-powered recommendation system to improve the energy efficiency of Google’s data centres in 2016.




How DeepMind’s AI algorithm saves energy


These are four key steps:


  • · Every five minutes, DeepMind’s cloud-based AI pulls a snapshot of the data centre cooling system from thousands of sensors (i.e. temperatures, power, pump speeds, setpoints, etc.)


  • · Then, this data from sensors is fed into DeepMind’s deep neural networks, which predict how different combinations of potential actions will affect future energy consumption.


  • · AI system then identifies which actions will minimise the energy consumption while satisfying a robust set of safety constraints.


  • · Those actions are sent back to the data centre, where the actions are verified by the local control system and then implemented.



According to DeepMind’s website, DeepMind’s AI- system have successfully managed to reduce the amount of energy Google used to cool its data centres by 40%.



Since 2016, google data centre operators were responsible for acting on the recommendations given by DeepMind’s AI-powered recommendation system. But now Google has allowed the data centre operators to take a back seat, giving the AI administrative level control over google servers to optimize the energy usage at the data centers.



On one hand, it is amazing to see DeepMind’s AI system delivering energy savings in multiple Google data centres. But on the other hand, people like Elon Musk and Sam Harris have been continuously warning us about the dangers of giving AI too much control.



  • What if DeepMind’s AI system already having the administrative level control over google servers manages to take over the whole google system?


  • What if DeepMind’s AI can access our data collected by google ecosystem and use that against us?



References:

https://deepmind.com/blog/article/deepmind-ai-reduces-google-data-centre-cooling-bill-40

https://deepmind.com/blog/article/safety-first-ai-autonomous-data-centre-cooling-and-industrial-control

https://techwireasia.com/2020/05/has-google-cracked-the-data-centre-cooling-problem-with-ai/

https://www.forbes.com/sites/samshead/2018/08/18/google-trusts-deepmind-ai-to-manage-data-centre-cooling/#459dff0068bd

7 views0 comments