Bio
Emmanouel (Manos) Varvarigos received a Diploma in Electrical and Computer Engineering from the National Technical University of Athens (NTUA) in 1988, and the M.S. and Ph.D. degrees in Electrical Engineering and Computer Science from the Massachusetts Institute of Technology (MIT) in 1990 and 1992, respectively. He is currently the Vice Rector for Research, Innovation, and Extroversion of the National Technical University of Athens. He has held faculty positions at the University of California, Santa Barbara (1992-1998, ECE dept, as an Assistant and later an Associate Professor with tenure), Delft University of Technology, the Netherlands (1998-2000, ECE dept, as an Associate Professor), University of Patras, Greece (2000-June 2015, CEID, as a Professor), and National Technical University of Athens (June 2015-now, ECE dept, as a Professor). During Jan 2017 - Dec 2018 he was also a Professor and Head of the Electrical and Computer Systems Engineering Department at Monash University, on leave of absence from NTUA. He has served as the Director of the Hardware, Communications and Networking Division of CEID at the University of Patras and the Director of the Communications, Electronics, Information Systems and Networking Division of ECE at the National Technical University of Athens. From 2003-2017, he was also the Scientific Director of the Greek School Network and Network Technologies Division of the Computer Technology Institute “Diophantus” (CTI), which through its involvement in pioneering research and development projects, has a major role in the development of network technologies and telematic services in Greece, and is responsible for development and operation of the Greek School Network, the largest public Network in Greece.
Professor Varvarigos has served in the organizing and program committees of over 150 international conferences and in national committees. He has participated in over 40 USA, Australian, and EU-funded research projects and in many national research projects, and has served as the consortium coordinator in several of them. He has also worked as a researcher at Bell Communications Research, and has consulted with several companies in the US and in Europe. He has over 480 publications in international journals and conferences. His research activities are in the areas of optical networking, edge and cloud computing, network algorithms and protocols, optical interconnects for Data Centers, smart energy grids, 5G networks, Machine Learning, and network services.
Abstract
Artificial intelligence workloads, from distributed training and latency-critical inference to emerging agentic applications, are reshaping the requirements placed on computing and network infrastructures. These workloads are distributed by nature, highly dynamic, and tightly coupled to the network interconnecting edge and cloud resources, so that computation placement and data movement can no longer be optimised in isolation. This talk examines how optimisation and learning can be combined to orchestrate such workloads across the edge-cloud continuum. We present exact and approximate formulations for the joint allocation of computing and network resources, with approximate dynamic programming and multi-agent rollout supporting offline optimisation and online decision making, and learning-based orchestration enabling autonomous adaptation under uncertainty and conflicting objectives. We further discuss market based coordination and congestion management under service-level guarantees, and the role of reconfigurable optical infrastructures that adapt to AI traffic patterns.
