首页> 外文会议>High-performance computing and networking >Modeling Speedup of SPMD Applications on the Intel Paragon: A Case Study
【24h】

Modeling Speedup of SPMD Applications on the Intel Paragon: A Case Study

机译:在Intel Paragon上对SPMD应用程序加速建模的案例研究

获取原文
获取原文并翻译 | 示例

摘要

Interconnection networks with nearly distance independent communication latency are a key feature of recent architectures. However, shared resources such as network channels can become bottlenecks that degrade performance and limit workload scalability. In this paper, the problem of network contention in medium scale multi-computers is addressed. A queueing network model that predicts application speedup on the Paragon is presented. The model has been experimentally validated under the SUNMOS operating system. The experimental and modeling results are in good agreement and suggest ways to avoid internal network contention of communication intensive applications on the Paragon.
机译:具有近乎距离无关的通信等待时间的互连网络是最新体系结构的关键功能。但是,诸如网络通道之类的共享资源可能会成为瓶颈,从而降低性能并限制工作负载的可伸缩性。本文解决了中型多计算机网络竞争的问题。提出了一个可预测Paragon上应用程序加速的排队网络模型。该模型已在SUNMOS操作系统下经过实验验证。实验和建模结果吻合良好,并提出了避免Paragon上通信密集型应用程序内部网络争用的方法。

著录项

  • 来源
  • 会议地点 Milan(IT);Milan(IT)
  • 作者

    E. Smirni; E. Rosti;

  • 作者单位

    Department of Computer Science, Vanderibilt Univeristy Nashville TN 37235, USA;

    Dipartimento di Scienze dell'Informazione Universita di Milano, Italy;

  • 会议组织
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类 TQ4;
  • 关键词

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号