...
首页> 外文期刊>Parallel Processing Letters >GUEST EDITOR'S NOTE: LARGE-SCALE PARALLEL PROCESSING
【24h】

GUEST EDITOR'S NOTE: LARGE-SCALE PARALLEL PROCESSING

机译:来宾编辑说明:大型并行处理

获取原文
获取原文并翻译 | 示例
           

摘要

The last decade has seen tremendous growth in the use of large scale parallel sys-ntems, containing thousands of processing nodes. Architecturally, these systems arenquite diverse, with different node processors, memory systems, and communicationnmodes. As a result, new challenges are constantly arising for researchers to find waysnof efficiently exploiting the resources of large scale systems. We are still far fromnthe point of achieving a parallel software development ecosystem that yields highnperformance and portability across a broad range of systems. The papers in thisnissue can be categorized as approaching the problem from two ends of the spectrum.nOn the one hand, we seek to find ways to broadly model system performance andnadaptively map applications for greater efficiency. On the other hand, we still neednto explore the implications of specific architectural configurations, in the context ofnparticular classes of applications, to better understand the ways in which we canntake greatest advantage of them.
机译:在过去的十年中,包含数千个处理节点的大规模并行系统的使用取得了巨大的增长。在体系结构上,这些系统完全不同,具有不同的节点处理器,内存系统和通信模式。结果,研究人员不断寻找新的挑战,以寻找有效利用大规模系统资源的途径。我们离实现并行软件开发生态系统还差得很远,该软件开发生态系统可在广泛的系统上实现高性能和可移植性。可以从频谱的两端来解决该问题。一方面,我们寻求找到广泛的系统性能建模方法,并自适应地映射应用程序以提高效率。另一方面,在特定的应用程序类别中,我们仍然需要探索特定体系结构配置的含义,以更好地理解我们最大程度地利用它们的方式。

著录项

  • 来源
    《Parallel Processing Letters》 |2011年第3期|p.275-277|共3页
  • 作者单位

    Department of Computer Science, University of Massachusetts, Amherst, MA 01003, USA;

    Pacific Northwest National Laboratory, Richland, WA 99353, USA;

    IBM Austin Research Laboratory, 11501 Burnet Road, Austin, TX 78726, USA;

  • 收录信息
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类
  • 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