Study on Power Demand Estimation Methods for Computing Power Infrastructure

Data centers, as the infrastructure for the development of the digital economy, have entered a fast-track phase in recent years. Currently, artificial intelligence is undergoing transformative changes, and the power demand brought about by the development of artificial intelligence poses challenges to society. The power consumption of computing power infrastructure has become a concern for all sectors of society. This paper analyzes the factors affecting the construction of two power demand estimation methods based on rack data and computing power data, proposes the future power demand of China’s computing power infrastructure through specific data calculations, and finally offers relevant suggestions.