Public Utilities

Machine Learning Targets Data-Centre Energy and Cooling Demand

Machine learning is being used to improve data-centre scheduling and cooling efficiency, including a reported 40% reduction in Google cooling energy.

A report on data-centre energy demand describes the role of AI and machine learning in improving the efficiency of computing infrastructure. It notes that machine learning can schedule work, manage shared hardware, and help debug cloud software so operators obtain more computing capacity from existing machines. The report also cites Google’s use of DeepMind machine learning for data-centre cooling, with a reported 40% reduction in cooling energy. These developments matter to public utilities because data centers are major electricity consumers and their cooling systems can affect water demand. More efficient computing and cooling can reduce pressure on utility networks while helping operators accommodate growing AI workloads. The example shows that AI can be both a driver of infrastructure demand and a tool for reducing the resources required to operate that infrastructure.

Original source: Progressive Robot

Related RAQI solutions

In-depth analysis on RAQI Insight

Work with RAQI