DOI: 10.5176/978-981-08-7656-2 A-51
Authors: Constantin Timm, Andrej Gelenberg, Peter Marwedel, Frank Weichert
Abstract: Nowadays, General Purpose Computing on GPUs (GPGPU) accelerates many industrial and scientific applications in the high performance computing (HPC) domain. Recently, GPU vendors, such as Nvidia and AMD, promoted the utilization of GPUs in embedded systems. First vendors in the automotive sector will use the computation power of GPUs for their driver assistance systems. In these systems energy constraints are omnipresent. This paper firstly discusses the theoretical background of an energy aware embedded system design including a GPGPUcapable
graphics chip. In order to support these theoretical considerations, secondly an energy and runtime evaluation of a
low power GPU/CPU system is presented. We demonstrate that a profitable GPU integration, seen from an energy perspective, strongly depends on the structure and the features of an application such as a high parallelizability and the utilization level of the graphics card. The evaluation of several real world benchmarks shows that increasing the system’s power consumption by integrating a GPU can lead to a reduced overall energy consumption of a system.