Implementing a streaming application on a processor array: A case study on the Epiphany architecture

This paper reports on a case study in which an at- size application is ported onto a commercially available processor array. Its purpose is threefold: (1) Determine the suitability of processor arrays for this kind of application. (2) Develop a runtime software infrastructure that supports streaming applications on processor arrays. (3) Gather data and insights into the resulting system performance and the factors that affect it. In this study, we port a video decoder onto the Parallella platform, a small credit-card sized board that includes a 16 core Epiphany processor array by Adapteva, along with a dual-core ARM processor, some programmable logic, and shared DRAM. We discuss the infrastructure required to execute the application on that platform and how the Epiphany architecture facilitated the construction of lockless distributed inter-core communication. In so doing, we build in part on a previous study on a much smaller example in [1].