多GPU系统虚实地址转换架构研究  

Research on virtual-physical address translation architectures of multi-GPU system

作　　者：魏金晖 李晨[1] 鲁建壮[1] WEI Jin-hui;LI Chen;LU Jian-zhuang(College of Computer Science and Technology,National University of Defense Technology,Changsha 410073,China)

机构地区：[1]国防科技大学计算机学院,湖南长沙410073

出　　处：《计算机工程与科学》2021年第2期228-234,共7页Computer Engineering & Science

基　　金：国防科技大学科研计划(ZK20-04);重点实验室基金(6142110180102);湖南省科技计划(2018XK2102);国家重点研发计划(2018YFB0204301)。

摘　　要：近年来,随着大数据的发展,GPU应用的数据集规模急剧增加,这对GPU的处理能力提出了挑战。由于摩尔定律即将达到极限,提升单一GPU的性能变得越发困难,而多GPU系统通过提升GPU处理器级的并行性,成为应对该挑战的一种解决方案。GPU制造商对内存虚拟化的支持进一步简化了多GPU系统的编程,提升了资源利用率。内存虚拟化需要地址转换的支持,而地址转换的开销对系统性能具有重要影响。研究了多GPU系统中2种常见的地址转换架构,即分布式地址转换架构和集中式地址转换架构,通过模拟实验对2种架构进行了深度分析和比较,在此基础上提出了优化地址转换设计的建议。In recent years,with the development of big data,the dataset size of GPU applications has increased significantly,which raises challenges for current GPUs.However,as Moore's Law reaches its limit,it is not easy to improve the performance of single GPU any further;Instead,multi-GPU systems have been shown to be an effective solution due to its GPU processor-level parallelism.The support for memory virtualization in multi-GPU systems further simplifies the programming and improves the resource utilization.Memory virtualization requires the support for address translation,and the overhead of address translation has important impact on system’s performance.This paper studies two common address translation architectures in multi-GPU systems,that is,distributed address translation architecture and centralized address translation architecture.Through simulation experiments,this paper ana-lyzes and compares the advantages and drawbacks of two address translation architectures in-depth.On this basis,this paper proposes optimization suggestions for address translation in multi-GPU systems.

关 键 词：多GPU系统 内存虚拟化 地址转换架构 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]