DELFT UNIVERSITY OF TECHNOLOGY REPORT 09-12 Fast Iterative Solution of Large Sparse Linear Systems on Geographically Separated Clusters
نویسندگان
چکیده
Parallel asynchronous iterative algorithms exhibit features that are extremely well–suited for Grid computing, such as lack of synchronisation points. Unfortunately, they also suffer from slow convergence rates. In this paper we propose using asynchronous methods as a coarse–grain preconditioner in a flexible iterative method, where the preconditioner is allowed to change in each iteration step. A full implementation of the algorithm is presented using Grid middleware that allows for both synchronous and asynchronous communication. Advantages and disadvantages of the approach are discussed. Numerical experiments on heterogeneous computing hardware demonstrate the effectiveness of the proposed algorithm on Grid computers, with application to large 2D and 3D bubbly flow problems.
منابع مشابه
Fast iterative solution of large sparse linear systems on geographically separated clusters
Parallel asynchronous iterative algorithms exhibit features that are extremely well-suited for Grid computing, such as lack of synchronization points. Unfortunately, they also suffer from slow convergence rates. In this paper we propose using asynchronous methods as a coarse-grained preconditioner in a flexible iterative method, where the preconditioner is allowed to change in each iteration st...
متن کاملDELFT UNIVERSITY OF TECHNOLOGY REPORT 08-08 Solving Large Sparse Linear Systems Efficiently on Grid Computers using an Asynchronous Iterative Method as a Preconditioner
In this paper we describe an efficient iterative algorithm for solving large sparse linear systems on Grid computers and review some of its advantages and disadvantages. The algorithm is a combination of a synchronous flexible outer iterative method and a coarse–grain asynchronous inner iterative method as a preconditioner. We present results of a complete implementation using mature Grid middl...
متن کاملDELFT UNIVERSITY OF TECHNOLOGY REPORT 08-12 Parallel Scientific Computing on Loosely Coupled Networks of Computers
Efficiently solving large sparse linear systems on loosely coupled networks of computers is a rich and vibrant field of research. The induced heterogeneity and volatile nature of the aggregated computational resources present numerous algorithmic challenges. Designing efficient numerical algorithms for said systems is a complex process that brings together many different scientific disciplines....
متن کاملDELFT UNIVERSITY OF TECHNOLOGY REPORT 02-12 Some Numerical Aspects for Solving Sparse Large Linear Systems Derived from the Helmholtz Equation
In this report, several numerical aspects and difficulties for solving a linear system derived from the time-harmonic wave equations are overviewed. The presentation begins with the derivation of the governing equation for waves propagating in general inhomogeneous media. Due to the need of numerical solutions, various discretizations based on finite difference are discussed. Some numerical met...
متن کاملUnit commitment by a fast and new analytical non-iterative method using IPPD table and “λ-logic” algorithm
Many different methods have been presented to solve unit commitment (UC) problem in literature with different advantages and disadvantages. The need for multiple runs, huge computational burden and time, and poor convergence are some of the disadvantages, where are especially considerable in large scale systems. In this paper, a new analytical and non-iterative method is presented to solve UC p...
متن کامل