MIAI-DAYS : Tuesday May 4th

Pour une IA durable / AI and sustainability

Vers un apprentissage frugal

Environmental impacts of ICT : how can we walk on the path of digital frugality?

Laurent Lefevre is a permanent researcher in computer science at Inria (the French Institute for Research in Computer Science and Control). He is a member of the Algorithms and Software Architectures for Distributed and HPC Platforms (Avalon) team from the LIP laboratory in Ecole Normale Superieure of Lyon, France. He has organized several conferences in dsitributed networking and computing and he has been member of several program committees. He has co-authored more than 100 papers published in refereed journals and conference proceedings. His interests include energy efficiency, environmental impacts and performance of large-scale distributed computing and networking infrastructures, high performance networks protocols and services. Laurent Lefevre is responsible of the Grid’5000 (now SILECS) Lyon site and member of the Grid’5000 (now SILECS) executive board. Laurent Lefevre is co-leader of the French GDS CNRS EcoInfo group which deals with the eco-responsibility and full life cycle impacts of ICT.

Summary : Information and Communication Technology (ICT) continue to impact the environment through multiple factors. This talk will overview main impacts and propose through measurement, modeling and several research approaches to explore how we could operate reduction of such impact at large scale.

Comment mesurer l’impact du Numérique dans un projet ?

Thierry Leboucq is the chairman of GREENSPECTOR, which he founded in 2010. Author and speaker, he is extremely active and plays a pioneering role in the green information technology sector in France. He was a major contributor in France to software ecodesign concerns. He has led several R & D projects on green software and several initiatives on measurement and green labeling for Web, mobile and IoT applications.

Summury : How to measure the impact of digital in a project?
To respect a level of sobriety for an application, it is essential to manage the consumption of the digital service throughout a project, both in its
design and manufacturing phase but also in its day-to-day production. Piloting by measurement will make it possible to observe deviations, gains
and consumption hotspots, but also to set goals and validate this requirement for sobriety. What operational and environmental indicators,
how to measure them, how to set thresholds?

Artificial intelligence and ressource optimization

Francis Bach - Researcher at Inria, leading since 2011 the machine learning team which is part of the Computer Science department at Ecole Normale Supérieure. Francis Bach is primarily interested in machine learning, and especially in sparse methods, kernel-based learning, large-scale optimization, computer vision and signal processing.

He obtained in 2009 a Starting Grant and in 2016 a Consolidator Grant from the European Research Council, and received the Inria young researcher prize in 2012, the ICML test-of-time award in 2014 and 2019, as well as the Lagrange prize in continuous optimization in 2018, and the Jean-Jacques Moreau prize in 2019.

He was elected in 2020 at the French Academy of Sciences. In 2015, he was program co-chair of the International Conference in Machine learning (ICML), and general chair in 2018; he is now co-editor-in-chief of the Journal of Machine Learning Research.

Summury : Ressource optimization has always been a key driver in computer science researcher. The widespread use of artificial intelligence algorithms poses new challenges in the context of climate change. In this presentation, some of these new research questions will be presented.

Hardware architectures for low power AI

Bain-inspired strategies for minimizing power consumption in spiking neural network hardware

Giacomo Indiveri - Professor at the Faculty of Science of the University of Zurich and at Department of Information Technology and Electrical Engineering of ETH Zurich, Switzerland. 

His latest research interests lie in the study of spike-based learning mechanisms and recurrent networks of biologically plausible neurons, and in their integration in real-time closed-loop sensory-motor systems designed using analog/digital circuits and emerging memory technologies. He is a recipient of three European Research Council grants and a senior member of the IEEE society.

Neuromorphic Intelligence (NI) hardware systems implement the principles of computation observed in the nervous system by exploiting the physics of their electronic devices to directly
emulate the biophysics of real neurons and synapses. In this lecture he will present examples of NI circuits, and demonstrate applications of NI processing systems to extreme-edge use cases, that require low power, local processing of the sensed data.

How to design a power frugal hardware for AI – the bio-inspiration way

Alexandre Valentian – CEA LETI
Head of System-on-Chip and Advanced Technologies laboratory, CEA Grenoble
Head of the chaire MIAI : Hardware for spike-coded neural networks exploiting hybrid CMOS non-volatile technologies