Machine Learning Applications for Data Center Optimization

editor score
Editor Score:
55
editor score
Quality:
60
editor score
Usability:
20
editor score
Reliability:
80
Asset Type
License
Free
Expiration
Tuesday, May 28, 2024
Geography
map
North America Central America South America Oceania Scandinavia Europe Africa Middle East Asia
Publication year
2014
Industry
IT

The modern data center (DC) is a complex interaction of multiple mechanical, electrical and controls systems. The sheer number of possible operating configurations and nonlinear interdependencies make it difficult to understand and optimize energy efficiency. Google developed a neural network framework that learns from actual operations data to model plant performance, it is a model that has been extensively tested and validated at Google DCs. The results demonstrate machine learning is an effective way of leveraging existing sensor data to model DC performance and improve energy efficiency.

Yegii expert:
Author:
Jim Gao
Time horizon:
2-5 Years
Yegii Reviews: