An evaluation of an energy efficient many-core SoC with parallelized face detection
Hiroyuki Usui, Jun Tanabe, Toru Sano, Hui Xu, Takashi Miyamori · 2014
New applications such as image recognition and augmented reality (AR) have become into practical on embedded systems. For these applications, we have developed a many-core SoC that includes two many-core clusters with 32 energy efficient processor cores connected by a low latency tree-based NoC. In this paper, we evaluate performance of many-core SoC by face detection as an example of real image recognition applications and discuss two parallelized implementations on the many-core clusters. By keeping balance of workloads on the cores, the performance scales up to 64 cores and the SoC consumes only 2.21W. The energy efficiency is several tens of times better than that of a high performance desk-top quad-core processor.