ParallelME: A parallel mobile engine to explore heterogeneity in mobile computing architectures

Abstract

Following the evolution of desktops, mobile architectures are currently witnessing growth in processing power and complexity with the addition of different processing units like multi-core CPUs and GPUs. To facilitate programming and coordinating resource usage in these heterogeneous architectures, we present ParallelME, a Parallel Mobile Engine designed to explore heterogeneity in mobile computing architectures. ParallelME provides a high-level library with a friendly programming language abstraction for developers, facilitating the programming of operations that can be translated into low-level parallel tasks. Additionally, these tasks are coordinated by a runtime framework, which is responsible for scheduling and controlling the execution on the low-level platform. ParallelME’s purpose is to explore parallelism with the benefit of not changing the programming model, through a simple programming language abstraction that is similar to sequential programming. We performed a comparative analysis of execution time, memory and power consumption between ParallelME, OpenCL and RenderScript using an image processing application. ParallelME greatly increases application performance with reasonable memory and energy consumption.