U.S. Spain Cooperative Research: Parallel Solutions of Linear Systems

Szyld该奖项将支持坦普尔大学的丹尼尔Szyld和西班牙瓦伦西亚理工大学的Rafael Bru之间的合作研究。 拟议的研究将集中在大型线性代数方程组的并行解决方案。 特别是，异步迭代方法和预处理Krylov型方法，奇异和非奇异系统，将进行研究。 合作的主要好处是加入互补的技能和兼容的研究观点。 Bru教授关于非负矩阵乘积的收敛性的工作与Szyld关于分析两阶段迭代法和异步方法的收敛性的工作有关，并且已经导致合作者或他们的学生联合出版。 建议的研究包括理论，算法和实验元素。 它将探讨并行块和异步迭代方法的收敛特性和实现细节。 将进行数值实验。

Szyld This award will support collaborative research between Daniel Szyld of Temple University, and Rafael Bru, Polytechnic University of Valencia, Spain. The proposed study will focus on the parallel solution of large systems of linear algebraic equations. In particular, asynchronous iterative methods and preconditioned Krylov-type methods, for both singular and nonsingular systems, will be studied. The primary benefit of the collaboration is the joining of complementary skills and compatible research views. Professor Bru's work on the convergence of the product of nonnegative matrices is relevant to Szyld's work on analyzing the convergence of two-stage iterative methods and on asynchronous methods, and has already led to joint publications by the collaborators or their students. The proposed research includes theoretical, algorithmic, and experimental elements. It will explore convergence properties and implementation details of parallel block and asynchronous iterative methods. Numerical experiments will be performed.