The convergence variability of parallel iterative solvers
The convergence variability of parallel iterative solvers
复制标题
DOI：
10.1108/02644400610644522
复制
发表时间：
2006-02
影响因子：
1.6
通讯作者：
I. Smith;L. Margetts
中科院分区：
文献类型：
--
作者：
I. Smith;L. Margetts
目的-调查一个众所周知的现象与一系列的并行迭代求解器的原因-实现收敛所需的迭代次数的变化。设计/方法/途径-结论是基于使用并行计算机进行广泛的实验。最近发表的作品也被用来提供迭代计数的变化的其他例子。结果-并行化，元素由元素迭代求解器是由数值精度和roundeness.研究限制/影响-理论检查的现象可能会带来光的方法，其中迭代计数可以限制到变量范围的低端-从而减少解决方案的时间。实际影响-作者认为，在迭代计数的变化描述的元素的元素的方法提出了没有真实的困难的工程分析。独创性/价值-本文给出.
Purpose – To investigate the cause of a well‐known phenomenon associated with a range of parallel iterative solvers – the variability in the number of iterations required to achieve convergence.Design/methodology/approach – The conclusions are based on extensive experiments undertaken using parallel computers. Recently published works are also used to provide additional examples of variability in iteration count.Findings – The variability of iteration counts experienced by parallelised, element‐by‐element iterative solvers is caused by numerical precision and roundoff.Research limitations/implications – A theoretical examination of the phenomenon may bring to light a methodology in which the iteration count could be limited to the lower end of the variable range – thus reducing solution times.Practical implications – The authors believe that the variability in iteration count described for element‐by‐element methods presents no real difficulty to the engineering analyst.Originality/value – The paper gives...