Power & Thermal Efficient Architectures for Edge and IoT

Abstract

ABSTRACT The rapid proliferation of Edge and Internet-of-Things (IoT) devices has posed significant challenges with respect to power consumption and thermal management. As the demand for real-time processing increases, systems are being pushed to operate closer to data sources, necessitating new architectural strategies to optimise energy delivery and heat dissipation efficiently. This paper reviews recent developments in power and thermal efficient architectures specifically designed for Edge and IoT environments. We discuss both hardware and software level solutions, including energy harvesting, low power processors, dynamic voltage/frequency scaling (DVFS), heterogeneous architectures, and thermal aware scheduling. A comparative analysis highlights strengths and weaknesses of various approaches. Finally, future directions in designing sustainable, efficient computing at the edge are suggested. The findings in this review aim to support researchers and practitioners in addressing critical constraints in contemporary IoT deployments.

KEYWORDS: Edge Computing, Internet of Things (IoT), Power Efficiency, Thermal Management, Heterogeneous Architecture, Low-Power Hardware, DVFS, Energy Harvesting