The 2005 International Conference on Preconditioning Techniques for Large Sparse Matrix Problems in Industrial Applications; May 19-21, 2005; Atlanta, GA
The 2005 International Conference on Preconditioning Techniques for Large Sparse Matrix Problems in Industrial Applications; May 19-21, 2005; Atlanta, GA
批准号:
0435964
负责人:
Michele Benzi
金额:
$1.13万
依托单位:
依托单位国家:
美国
项目类别:
Standard Grant
财政年份:
2004
资助国家:
美国
项目状态:
已结题
起止时间:
2004-09-01 至 2005-08-31
中文摘要
在许多大规模的科学应用和工业数值模拟中,最内部的计算内核通常是一个大的稀疏矩阵问题,这通常消耗了模拟所需的总体计算时间的很大一部分。许多矩阵问题是线性方程组的形式,尽管其他类型的问题,如特征值计算,也可能发生。如果这些稀疏矩阵求解器的性能能够得到改善,将会产生巨大的影响。求解大型稀疏矩阵方程的传统方法是使用直接法。这种方法在工业中通常是首选的,因为直接求解器对于中等规模的问题是强大和有效的。然而,前所未有的技术进步速度导致要处理的矩阵的大小急剧增长。例如,三维模拟的存储需求使得直接方法非常昂贵。迭代技术是唯一可行的替代方法。不幸的是,迭代方法缺乏直接方法的鲁棒性。当矩阵条件非常恶劣时,它们往往会失败。这就是当前对大型稀疏矩阵问题的预处理技术产生巨大兴趣的原因:预处理显著地提高了迭代算法的鲁棒性和性能。尽管在过去的几年里,预处理技术取得了很大的进步,但迭代求解器仍然不总是完全可靠和有效的。提议的会议是关于预处理的一系列两年一次的会议中的第四次会议,旨在解决与大规模实际应用和工业环境中一般稀疏矩阵问题解决相关的复杂问题。应用科学家和工业用户感兴趣的与稀疏矩阵软件相关的问题通常与学术界关注的问题大不相同。例如,对于应用科学家或工业用户来说,提高健壮性可能比找到一种能够获得速度的方法重要得多。内存使用也是一个重要的考虑因素,但在稀疏矩阵求解的学术研究中很少考虑。最后一个例子,在应用中解决的线性系统几乎总是一些非线性迭代(例如,牛顿)或优化循环的一部分。重要的是要考虑线性和非线性部分之间的耦合,而不是只关注线性系统。本次会议的主讲人将讨论稀疏矩阵问题预处理技术领域的一些最新进展。会议将允许来自学术界、工业界和政府实验室的参与者交流该领域的研究成果,并根据新兴范式(如并行处理和面向对象编程)探索可能的新方向。
英文摘要
The innermost computational kernel of many large-scale scientific applications andindustrial numerical simulations is often a large sparse matrix problem, which typicallyconsumes a significant portion of the overall computational time required by the simulation.Many of the matrix problems are in the form of systems of linear equations, although othertypes of problems, such as eigenvalue calculations, can occur too. A tremendous impactwill be made if the performance of these sparse matrix solvers can be improved.A traditional approach to solving large sparse matrix equations is to use directmethods. This approach is often preferred in industry because direct solvers arerobust and effective for moderate size problems. However, the unprecedented paceof the advance in technology has led to a dramatic growth in the size of the matricesto be handled. For example, the storage requirement for three-dimensional simulations makesdirect methods prohibitively expensive. Iterative techniques are the only viable alternative.Unfortunately, iterative methods lack the robustness of direct methods. They often failwhen the matrix is very ill-conditioned. This is the reason for the tremendous currentinterest in preconditioning techniques for large sparse matrix problems: preconditionersdramatically improve the robusteness and performance of iterative algorithms. In spite of much progress on preconditioning techniques over the last few years,iterative solvers are still not always completely reliable and efficient. The proposedconference, the fourth in a series of biannual meetings on preconditioning, intends toaddress the complex issues related to the solution of general sparse matrix problemsin large-scale real applications and in industrial settings. The issues related to sparse matrix software that are of interest to application scientists and industrialusers are often fairly different from those on which the academic community is focused.For example, for an application scientist or an industrial user, improving robustnessmay be far more important than finding a method that would gain speed. Memory usage isalso an important consideration, but is seldom accounted for in academic research onsparse matrix solvers. As a last example, linear systems solved in applications arealmost always part of some nonlinear iteration (e.g., Newton) or optimization loop.It is important to consider the coupling between the linear and nonlinear parts, insteadof focusing on the linear systems alone. The speakers of this conference will discusssome of the latest developments in the field of preconditioning techniques for sparsematrix problems. The conference will allow participants from academia, industry andgovernment labs to exchange findings in this area and to explore possible new directionsin light of emerging paradigms, such as parallel processing and object-oriented programming.
期刊论文(0)
专著(0)
科研奖励(0)
会议论文
Generalized Matrix Functions: Theory, Algorithms, and Applications
-
批准号:1719578
-
项目类别:Standard Grant
-
资助金额:$25.0万
-
财政年份:2017
-
负责人:Michele Benzi
-
依托单位:
Numerical Methods for Graph and Network Analysis
-
批准号:1418889
-
项目类别:Standard Grant
-
资助金额:$18.0万
-
财政年份:2014
-
负责人:Michele Benzi
-
依托单位:
Numerical Linear Algebra Tools for the Analysis of Complex Networks
-
批准号:1115692
-
项目类别:Standard Grant
-
资助金额:$30.3万
-
财政年份:2011
-
负责人:Michele Benzi
-
依托单位:
Approximation of Matrix Functions: Theory, Algorithms, and Software
-
批准号:0810862
-
项目类别:Standard Grant
-
资助金额:$22.95万
-
财政年份:2008
-
负责人:Michele Benzi
-
依托单位:
Scalable Iterative Solution of Large Linear Systems with Applications in Fluid Dynamics, Radiation Transport and Markov Chains
-
批准号:0511336
-
项目类别:Standard Grant
-
资助金额:$0.0万
-
财政年份:2005
-
负责人:Michele Benzi
-
依托单位:
Development, Analysis, and Implementation of Robust Algebraic Preconditioners for Sparse Linear Systems
-
批准号:0207599
-
项目类别:Standard Grant
-
资助金额:$13.82万
-
财政年份:2002
-
负责人:Michele Benzi
-
依托单位:
海外基金