In this infographic we explore research on federated learning by Xinchi Qui et al. (2021). Artificial intelligence needs to be trained through machine learning to gather insights from data and automate tasks. This data is provided by centralised data centres. Centralised data centres have high carbon emissions and so researchers have been exploring federated learning as an alternative with reduced emissions! Federated learning trains AI across a range of devices, rather than one data centre. Find out more in our infographic!
Sources:
Qui, X., Parcollet, T., Fernandez-Marquez, J., Gusmao, P., Beutel, D.,Topal, T., Mathur, A. and Lane, N., (2021), 'A First Look into the Carbon Footprint of Federated Learning', arXiv
Qui, Xinchi, (2021, 1 July), 'What is the Carbon Footprint of Federated Learning?', Flower
CO2 in Federated Learning, Carbon Calculator, CaMLSys https://mlsys.cst.cam.ac.uk/carbon_fl/