Résumé
Information and Communication Technologies (ICT) represent a significant share of global resource usage. Notably, the energy consumption of data centres has emerged as a critical concern, escalating rapidly, especially with the rise of generative Artificial Intelligence (AI) models. This surge in energy demand calls for efforts focused on measuring and reducing the energy and carbon footprints of components used in ICT service delivery. However, the variety and multitude of devices involved in these services make it challenging to accurately measure and evaluate these footprints.
In this paper, we propose an alternative approach to assess energy impacts by considering data centres for what they are: throughput computing systems offering ICT services. We define a simple experimental testbed and evaluate several machines with different hardware capacities. These machines serve two ICT services: a loop incrementing a counter until a defined limit, and an AI inference predicting the next token based on an input.
Our experiment highlights three main take-aways: (a) the CPU usage is a poor predictor of the power consumption, (b) the energy or CO2 quantity associated with a service visit highly depends on the total load the service is facing, and (c) modern machines do not yield better energy figures compared to older ones in all circumstances.