Journal of Parallel and Distributed Computing ( IF 3.4 ) Pub Date : 2019-12-26 , DOI: 10.1016/j.jpdc.2019.12.011 Aaron Stillmaker , Brent Bohnenstiehl , Lucas Stillmaker , Bevan Baas
Three parallel sorting applications and two list output protocols for the first phase of an external sort execute on a fine-grained many-core processor array that contains no algorithm-specific hardware acting as a co-processor with a variety of array sizes. Results are generated using a cycle-accurate model based on measured data from a fabricated many-core chip, and simulated for different processor array sizes. The data shows most energy efficient first-phase many-core sort requires over 65 lower energy than GNU C++ standard library sort performed on an Intel laptop-class processor and over 105 lower energy than a radix sort running on an Nvidia GPU. In addition, the highest first-phase throughput many-core sort is over 9.8 faster than the std::sort and over 14 faster than the radix sort. Both phases of a 10 GB external sort require 6.2 lower energytime energy delay product than the std::sort and over 13 lower energytime than the radix sort.
中文翻译:
细粒度的多核处理器阵列上的可扩展节能并行排序
外部排序的第一阶段的三个并行分类应用程序和两个列表输出协议在细粒度的多核处理器阵列上执行,该阵列不包含特定算法的硬件,充当具有各种阵列大小的协处理器。结果是使用周期精确的模型根据制造的多核芯片的测量数据生成的,并针对不同的处理器阵列大小进行了仿真。数据显示最节能的第一阶段多核排序需要超过65 与在英特尔笔记本电脑级处理器上执行的105种以上GNU C ++标准库排序相比,能耗更低比在Nvidia GPU上运行的基数排序低的能量。此外,最高的第一阶段吞吐量多核排序超过9.8 比std :: sort快14比基数排序快。10 GB外部排序的两个阶段都需要6.2 较低的能量时间能量延迟乘积比std :: sort超过13 较低的能量时间比基数排序。