A MapReduce framework implementation for Network-on-Chip platforms

Abstract

All facets of society are generating increasing amounts of data confirming the term big data for modern applications. The next generation of embedded systems will be dominated by such smart applications offering a wide range of communication services. Driven also by hardware changes and the adoption of the many-core architectural template, a better resource management scheme is required. MapReduce is a programming model capable of processing large data sets with a parallel distributed algorithm using a large number of processing nodes. In this paper, we present a MapReduce framework for an embedded many-core Network-on-Chip platform with distributed shared memory characteristics. The proposed framework, which supports bare-metal systems, provides a scalable solution for data processing in a many-core system, while fully utilizing the platform's characteristics and achieving application speedup.