计算机科学 ›› 2018, Vol. 45 ›› Issue (6A): 262-265.
罗殊彦,朱怡安,曾诚
LUO Shu-yan, ZHU Yi-an,ZENG Cheng
摘要： 随着嵌入式技术的不断发展,越来越多的平台采用异构多核处理器(Heterogeneous Multi-Processor Unit,HMPU)进行高性能计算,但多核处理器的核间通信效率严格地制约着系统的高性能计算能力。针对HMPU的核间通信性能难以量化的问题,提出了基于通信粒度、通信缓存和消息传输机制的阶段评价模型,并通过实验验证了这3个影响因子对不同阶段的核间通信性能的影响。此外,由于嵌入式系统环境多变、资源有限,使得静态通信策略对系统性能优化具有局限性。针对该问题,提出了基于系统内存约束、时间约束和性能目标的动态通信策略优化模型(Dynamic Communication Strategy Optimization Model,DCSOM)。通过实验证明:在数据量较小、通信周期较长的异构多核处理器中,DCSOM更具优越性。
中图分类号:
| [1]STANISIC L,THIBAULT S,LEGRAND A,et al.Modeling and Simulation of a Dynamic Task-Based Runtime System for Heterogeneous Multi-core Architectures[C]∥International Conference on Parallel Processing.2014:50-62.<br /> [2]MITTALS,VETTER J S.A Survey of CPU-GPU Heterogeneous Computing Techniques[J].ACM Computing Surveys,2015,47(4):1-35.<br /> [3]冯强,胡毅,于东,等.基于OMAP处理器的核间通信机制设计与实现[J].计算机工程,2014(4):281-286.<br /> [4]张攀勇,孟丹,霍志刚.多核环境下高效集合通信关键技术研究[J].计算机学报,2010(2):317-325.<br /> [5]韩乐,陈香兰,李曦.一种降低核间通信开销的调度算法[J].计算机系统应用,2014(9):65-71.<br /> [6]胡孔阳,韩琼磊.一种应用于高性能DSP的RapidIO控制器IP[J].中国集成电路,2015(Z1):35-38.<br /> [7]YANG L,LIU W,JIANG W,et al.Application Mapping and Scheduling for Network-on-Chip-Based Multiprocessor System-on-Chip With Fine-Grain Communication Optimization[J].IEEE Transactions on Very Large Scale Integration Systems,2016,24(10):3027-3040.<br /> [8]XUT C,LEPP NEN V.Cache- and Communication-aware Application Mapping for Shared-cache Multicore Processors[C]∥Uksim/amss European Symposium on Computer Modeling & Simulation.IEEE,2015:115-120.<br /> [9]CASSANOL,COZZI D,JUNGEWELTER D,et al.An inter- processor communication interface for data-flow centric heterogeneous embedded multiprocessor systems[C]∥IEEE International Conference on Design & Technology of Integrated Systems in Nanoscale Era.IEEE,2014:1-6.<br /> [10]KADJIH G E,OROU J B C,SANJUN M A F.Research on Performance Detection and Optimization Strategy for Multi-core Microprocessor[J].Electric Power Information & Communication Technology,2014,556-562(18):3408-3411.<br /> [11]KURIANG,SUN C,CHEN C H O,et al.Cross-layer Energy and Performance Evaluation of a Nanophotonic Manycore Processor System Using Real Application Workloads[C]∥International Parallel and Distributed Processing Symposium.IEEE,2012:1117-1130.<br /> [12]SRIKANTHANS,DWARKADAS S,SHEN K.Data sharing or resource contention:toward performance transparency on multicore systems[C]∥Usenix Conference on Usenix Technical Conference.USENIX Association,2015:529-540.<br /> [13]NAVIKHANF,GOVIL K.Cluster based Optimization Routing Strategy for Data Communication in Mobile Computing[J].International Journal of Computer Applications,2014,99(2):19-24. |
| [1] | 孟德龙,文敏华,韦建文,林新华. 神威太湖之光上OpenFOAM的移植与优化 Porting and Optimizing OpenFOAM on Sunway TaihuLight System 计算机科学, 2017, 44(10): 64-70. https://doi.org/10.11896/j.issn.1002-137X.2017.10.012 |
| [2] | 蒋建春，汪同庆. 一种异构多核处理器嵌入式实时操作系统构架设计 Architecture Design of Embedded Real-time Operating System for Heterogeneous Multi-core Processor 计算机科学, 2011, 38(6): 298-. |
| [3] | . 基于多核处理器并发计算软件构架设计与实现 计算机科学, 2008, 35(7): 283-285. |
|
||
首页