Before
Data centres currently are estimated to be using almost 200 terawatt hours (TWh) of electricity per year; that’s a staggering 1% of all global energy consumption.
After
Power usage effectiveness (PUE) is an important industry metric for measuring the energy efficiency of a data center. Optimising this can provide huge cost savings of up to 20% in this case
The How
- Using MAIO, collect operational data from the site
- Data engineering to clean and prepare the data sets
- Develop a time lagged model, predicting PUE into the future by approximately 1 hour.
- Using this model as a simulator, mathematical optimisation methods are developed
- Develop an intuitive UI for data centre engineers to generate optimal setpoint values that minimise PUE.
