Exploring the performance–power–energy balance of low-power multicore and manycore architectures for anomaly detection in remote sensing

Abstract

In this paper, we perform an experimental study of the interactions between execution time (i.e., performance), power, and energy that occur in modern low-power architectures when executing the RX algorithm for detecting anomalies in hyperspectral images (i.e., signatures which are spectrally different from their surrounding data). We believe this is important because, for airborne and spaceborne remote sensing missions, power and/or energy can be in practice as relevant as performance. In this sense, this paper investigates whether several recent low-power multithreaded architectures, from ARM and NVIDIA, can be a practical alternative in this domain to a standard high-performance multicore processor, using the RX anomaly detector as a case study.