EE Distinguished Speakers Seminar: Designing Computing Systems for the Internet of Things: Follow Your Brain!
The Internet of Things (IoT) has been hailed as the next frontier of innovation in which our everyday objects are connected in ways that improve the quality of our living environments, as well as industrial efficiency. As a result, the IoT concept is poised to reach 40 billion connected “things” (or devices) by 2025. However, major key challenges remain in achieving this potential due to the inherent resource-constrained nature of IoT systems, coupled with the computing power requirements of Big Data computing servers, which can result in degraded and unreliable behavior computing systems, or a global energy crisis when IoT is fully deployed. In this seminar, first, the challenges of ultra-low power design and communication in IoT devices will be presented. Then, possible design approaches for wearables and next-generation self-aware edge AI devices will be analyzed to successfully develop energy-efficient intelligent environments for the Big Data Era. As a result, it will be discussed how to optimize cloud computing server infrastructures to interact with edge devices by including a new three-dimensional (3D) computing many-core architecture that overcomes the prevailing worst-case power and cooling provisioning paradigm for centralized computing. This novel 3D server design includes a new system-level machine learning-based proactive approach for heat and energy management by capitalizing on the latest micro-scale two-phase liquid cooling technology. Finally, inspired by the mammalian central nervous system, it will be shown how we can integrate on this new 3D many-core a novel technology of on-chip microfluidic fuel cell networks to enable energy-scalability in future generations of servers for the IoT Era.
Zoom link: https://epfl.zoom.us/j/95573701198 / Id. 955 7370 1198