面向大数据应用挑战的超级计算机设计

On the challenge for supercomputer design in the big data era

在大数据时代,大数据处理给面向高性能计算设计的超级计算机带来计算、存储、通信、编程等软硬件体系结构方面的挑战.对此在分析目前主流学术界和工业界相关研究成果的基础上,介绍了天河二号超级计算机系统针对大数据处理应用所采用的新型异构多态体系结构、自主定制高速互连网络、混合层次式存储系统、混合计算框架等多种优化措施,为大数据时代的超级计算机设计提供有意义的参考.

Abstract： Because traditional supercomputer is designed for high-performance comput- ing, big data processing applications brings some software and hardware challenges including compute, storage, communication and programming. This paper introduces optimization methods of Tianhe-2 supercomputer system to process big data, such as a new heterogeneous polymorphic architecture, custom high-speed TH-Express 2＋ interconnec- tion network, hybrid hierarchical storage system and hybrid computing pattern framework.These efforts maybe make help for how to design supercomputers in the age of big data.