United States.
Google has announced that it has applied its DeepMind machine learning algorithms to its own data centers to reduce by up to 40% the amount of energy used for cooling.
One of the main sources of power consumption in the data center environment is cooling. However, Google said dynamic environments like data centers make it difficult to operate optimally for several reasons:
1. The equipment, the way it is operated, and the environment interact with each other in a complex, non-linear way. engineering based on traditional formula and human intuition often fail to capture these interactions.
2. The system cannot quickly adapt to internal or external changes (such as weather). This is because we cannot come up with rules for every operation scenario.
3. Each data center has a unique architecture and environment. A model with a custom configuration for one system may not be applicable to another. Therefore, an intelligent overall framework is needed to understand data center interactions.
To address this problem, Google began applying machine learning two years ago to operate its data centers more efficiently. And in recent months, DeepMind researchers began working with Google's data center team to improve the system's usefulness.
Using a system of neural networks trained in different operational scenarios and parameters within their data centers, the team created a more efficient and adaptable framework for understanding data center dynamics and optimizing efficiency.
This was achieved by adopting historical data that had already been collected by data sensors – such as temperature, power, pump speed, lockers, etc. – and using it to train neural networks.
Google said the machine learning system was able to consistently achieve a 40% reduction in the amount of energy used for cooling.
