From Edge to Extreme Edge: Exploring the Dimensions of AI Inference in IoT | Roberto Morabito
May 02, 2024 | 4:00 PM CET
About Topic:
This talk presents an in-depth exploration of the dual facets of AI inference within the IoT ecosystem: the established edge computing and the emerging "extreme edge", where TinyML is gaining prominence. We delve into how these two spheres interact and complement each other, driving advancements in AI-driven IoT applications. The discussion highlights how the proliferation of AI inference accelerators and varying software, hardware, and networking configurations are shaping the landscape of computational efficiency and real-time responsiveness. We will explore the development of frameworks that accommodate hardware diversity, ensuring optimized AI inference performance across a spectrum of devices. The seminar will also shed light on the strategies for seamless integration and coordination of devices at both the edge and the extreme edge. Emphasizing the importance of data-driven decision-making, we will discuss how these technologies enable robust AI capabilities in diverse applications. As a future perspective, we will briefly consider the potential role of emerging AI technologies, such as generative AI, in further enhancing the capabilities and applications of AI inference within IoT.
About Speaker:
Roberto Morabito is an Assistant Professor in the Communication Systems Department at EURECOM, France. Before joining EURECOM, he was a Senior Researcher at the Department of Computer Science, University of Helsinki. From May 2014 to April 2022, he was part of Ericsson Research Finland, contributing to the Cloud and System Platforms teams until March 2018, and subsequently to the IoT Technologies and Cyber-Physical Systems team. Between 2014 and 2018, Roberto was a Marie Skłodowska-Curie research fellow within the FP7 ITN project 'METRICS.' He earned his PhD in Networking Technology from Aalto University, Finland, in May 2019. From June 2019 to March 2021, he served as a Postdoctoral Researcher at the EDGE LAB, School of Electrical and Computer Engineering, Princeton University, USA. Roberto has also been a visiting researcher at INRIA Lille, France, the Technical University of Munich, Germany, and Yale University, USA. His work intersects IoT, Edge Computing, and Distributed AI, focusing on trade-offs in AI service provisioning and orchestration under computing and networking resource constraints. Recently, he started exploring the impact of generative AI, like Large Language Models, in these contexts. Previously, he investigated diverse topics in networked systems, including IoT protocols, virtualization, and service mesh technologies.