...
首页> 外文期刊>Transactions of the American nuclear society >Validation of Full-Domain Massively Parallel Transport Sweep Algorithms
【24h】

Validation of Full-Domain Massively Parallel Transport Sweep Algorithms

机译:全域大规模并行传输扫描算法的验证

获取原文
获取原文并翻译 | 示例

摘要

These computational results validate the predictions made by the sweep algorithm performance models. Codes based on sweep algorithms and domain decompositions implemented as described here and in the referenced publications will scale to millions of MPI ranks and potentially beyond. The strength of this study is the demonstration of the validity of sweep algorithm scaling predictions using two vastly different codes with markedly different domain decomposition and spatial discretization choices. Additional research into algorithms and strategies, as well as code improvements, will almost certainly result in improved efficiencies.
机译:这些计算结果验证了扫掠算法性能模型所做的预测。如此处和参考出版物中所述,基于扫描算法和域分解的代码将扩展到数百万个MPI等级,甚至可能更高。这项研究的优势在于证明了使用两种截然不同的域分解和空间离散化选择的代码,扫描算法缩放预测的有效性。对算法和策略的附加研究以及代码改进几乎可以肯定会导致效率提高。

著录项

  • 来源
    《Transactions of the American nuclear society》 |2014年第11期|699-702|共4页
  • 作者单位

    Department of Nuclear Engineering, Texas A&M University, College Station, TX 77843-3133;

    Lawrence Livermore National Laboratory, Livermore CA, 94551;

    Department of Nuclear Engineering, Texas A&M University, College Station, TX 77843-3133;

    Lawrence Livermore National Laboratory, Livermore CA, 94551;

    Lawrence Livermore National Laboratory, Livermore CA, 94551;

    Department of Nuclear Engineering, Texas A&M University, College Station, TX 77843-3133;

    Department of Computer Science and Computer Engineering, Texas A&M University, College Station, TX 77843-3112;

    Department of Computer Science and Computer Engineering, Texas A&M University, College Station, TX 77843-3112;

    Department of Computer Science and Computer Engineering, Texas A&M University, College Station, TX 77843-3112;

  • 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