AI (Artificial Intelligence) has been slowly spreading into various industrial sectors over the past few years, so it should come as no surprise that AI algorithms are now being deployed in Data Centres – most commonly utilised to optimise and monitor the intricate cooling systems.
The most apparent way that AI is already changing how operators are cooling their Data Centres is by significantly improving efficiency. AI neural networks can assess data from cooling systems, in real-time and are now able to not only make accurate predictions about how potential actions will affect the efficiency of those systems, but implement them too.
This level of meticulous system monitoring and real-time adjustment of settings is not something that would be feasible using manual monitoring, as it would prove to be an inefficient use of resources.
Of course, the improved efficiency in Data Centre cooling facilitated by AI isn’t just about changing the way the equipment is cooled, but also about making the entire process as efficient as possible. Given that the latest estimate suggest that 2% of greenhouses emissions worldwide are produced by Data Centres, making their energy consumption as efficient as possible isn’t just about saving money; it’s a wider corporate social responsibility.
Part of the Overall Implementation of AI in Data Centres
Cooling is only a small part of the larger picture of how AI is changing Data Centres with predictions that by 2020 over 30% of Data Centres that haven’t implemented AI or machine learning will no longer be operationally or commercially viable. Look at any trends and predictions in the world of Data Centres and it’s clear to see that AI will be taking control.
AI Cooling in Action
While for many the implementation of AI in Data Centres is still a dream rather than reality, Google is once again leading the way, having already entrusted the cooling of its Data Centres to AI since 2016. Originally only able to make recommendations, the automated system can now make tweaks on its own and implement those in real time.
Given the high level of efficiency Google’s Data Centres already displayed in most areas, any improvement in cooling would have been impressive, let alone the 40% reduction in energy usage the company achieved by implementing AI.
Since giving an AI control of its cooling systems, Google now owns 15 of the most energy efficient Data Centres in the world, proving that AI controlled cooling systems are not only viable but are already making a significant impact in the Data Centres where they have been deployed.
For more information about our award winning Data Centres, contact Sudlows today.