Curbing The Carbon Footprint Of Machine Learning

SustainML is a EU-funded project that seeks to create an innovative development framework to reduce the power consumption of Machine Learning applications. The project is coordinated by the Spanish middleware company eProsima and includes IBM, UPMEN, RPTU, l’Université de Copenhague, DFKI and Inria. The project focuses on quantifying the environmental impact of algorithms and the consequences of each decision taken throughout the ML life-cycle.