文档介绍：山东科学第 24 卷第 5 期 2011 年 10 月出版
SHANDONG SCIENCE Vol. 24 No. 5 Oct. 2011
文章编号: 1002-4026( 2011) 05-0056-07
基于异构模式的云计算关键技术研究
张庆科,杨波* ,王琳,陈贞翔
( 济南大学信息科学与工程学院,山东省网络环境智能计算技术重点实验室,山东济南 250022)
摘要: 结合云计算中 Map /Reduce 分布式编程技术引入了基于 CPU-GPU 异构混合并行编程模式,给出了该并行编程模式
的原理和实现过程。该模式通过采用 CUDA 多线程并行机制提高了大规模数据处理的效率。文中对比分析了云计算中
两种典型的分布式存储系统 GFS 和 HDFS,最后从宏观角度阐释了云计算虚拟化技术的三层部署架构和基本类型。
关键词: 云计算; 图形处理器( GPU) ; CUDA; 并行编程模型; 分布式存储; 虚拟化
中图分类号: TP393 文献标识码: A
Research on heterogeneous model based
key puting technologies
ZHANG Qing-ke,YANG Bo* ,WANG Lin,CHEN Zhen-xiang
( Shandong Provincial Key Laboratory work Based puting,
School of Information Science and Engineering,University of Jinan,Jinan 250022,China)
Abstract ∶ This paper presents a CPU-GPU heterogeneous parallel programming model based on the distributed
programming technology of puting,Map /Reduce. This paper also gives its principle and implementation
process. It improves the efficiency of large-scale data processing with the multi-thread parallel mechanism,CUDA. We
contrastively analyze two typical distributed storage systems,GFS and HDFS. We eventually present the three-layer
architecture and basic type of virtualization technology.
Key words ∶ puting; GPU; CUDA; parallel programming model; distributed storage; virtualization
计算机、互联网和通讯技术的快速发展使得网络对海量级数据存储能力和计算能力的需求日益提升。
云计算通过协同调度网络中现有的软硬件资源,实现了存储与计算服务模式的虚拟化和透明化,并以