Design and Implementation of Efficient Approach for Optimizing Energy Consumption for Data Centers

Abstract

Because of how big data works, the size and amount of data that must be stored and processed in hyperscale data centres (HDCs) grows as its use grows. Before starting work on an HDC, the financial plan needs to be looked at in depth in terms of the energy needs and the likely cost of energy. We give a framework based on power use effectiveness (PUE) for figuring out how much energy the world will need in the future. This framework takes into account all possible sources of heat and electrical load. The system uses both a physical model and a statistical framework to predict how much energy IT equipment and the data centre will use. The framework also includes a way to figure out the carbon emissions and electricity costs of the data centre. To figure out the annual PUE for sixty regions, we used hourly weather data as climate factors and a small set of energy attributes. For example, the Carbon Usage Effectiveness (CUE) and prices of electricity in India were looked at. Real-time data and experiments show that our method can accurately predict how much energy HDCs will use in total. This makes it possible to do an HDC feasibility study and fills a gap in HDC research in the Asia-Pacific region.