Development of Analysis Tools for Energy Efficiency Increase of Existing Data Centres
Ansis Avotiņš, Agris Ņikitenko, Armands Šenfelds, Janis Kikans, Andrejs Podgornovs, Mihails Sazonovs · 2022
Challenge of analysis and optimization method selection for data centre energy efficiency improvement has been addressed based on several pilot site data centres. Typically energy efficiency has been presented by single performance factor – power usage effectiveness. Full cycle of workflow and methods covering data collection, system modelling and complexity layer definition, data analysis and user feedback and visualization approaches have been discussed and proposed for in-depth analysis and impact on power usage effectiveness examination. Experimental plan is created and higher resolution sensors are selected to analyse thermodynamic behaviour of system and its elements, which will enable AI-based algorithm development for automated analysis, cooling regime optimisation and system behaviour prediction.