Accelerating Linux and Android applications on low-power devices through remote GPGPU offloading

摘要

Low-power devices are usually highly constrained in terms ofCPUcomputing power,memory, andrnGPGPU resources for real-time applications to run. In this paper, we describe RAPID, a completernframework suite for computation offloading to help low-powered devices overcome these limitations.rnRAPID supports CPU and GPGPU computation offloading on Linux and Android devices.rnMoreover, the framework implements lightweight secure data transmission of the offloadingrnoperations.We present the architecture of the framework, showing the integration of the CPUrnand GPGPU offloading modules. We show by extensive experiments that the overhead introducedrnby the security layer is negligible. We present the first benchmark results showing thatrnJava/Android GPGPU code offloading is possible. Finally, we show the adoption of the GPGPUrnoffloading into BioSurveillance, a commercial real-time face recognition application. The resultsrnshow that, thanks to RAPID, BioSurveillance is being successfully adapted to run on low-powerrndevices. The proposed framework is highly modular and exposes a rich application programmingrninterface to developers, making it highly versatile while hiding the complexity of the underlyingrnnetworking layer.