ALGORITHMS: Collaborative Research-SEIDD: Scalable Domain Decomposition Algorithms for Solving Parabolic Problems

算法：协作研究-SEIDD：解决抛物线问题的可扩展域分解算法

• 批准号: 0305355
• 负责人: Xian-He Sun
• 金额: $ 5.07万
• 依托单位: Illinois Institute of Technology
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2003
• 资助国家: 美国
• 起止时间: 2003-09-01 至 2005-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0305355&HistoricalAwards=false
• 关键词: ALGORITHMS; Collaborative; Research; SEIDD; Scalable

The Stabilized Explicit-Implicit Domain Decomposition (SEIDD) is a class of globally non-iterative and non-overlapping Domain Decomposition (DD) algorithms for solving parabolic equations on parallel computers. These algorithms do not use large amounts of memory and are inherently parallel, which makes them useful for large scale parallel processing.The SEIDD approach aims to provide algorithmic solutions that can better utilize parallel/distributed computing resources, thereby improving the accessibility of high performance computing power for researchers and educators from abroad range of science and engineering disciplines.Through integration with education and curriculum development, this research project promises to provide better education opportunities in high performance computing for a wide range of students while offering immediate training for computer science students in parallel algorithm design and programming.

稳定显隐区域分解（SEIDD）是一类在并行计算机上求解抛物型方程的全局非迭代、非重叠区域分解（DD）算法。这些算法不使用大量的存储器并且本质上是并行的，这使得它们对于大规模并行处理是有用的。SEIDD方法旨在提供可以更好地利用并行/分布式计算资源的算法解决方案，从而提高国外科学和工程学科研究人员和教育工作者获得高性能计算能力的能力。通过与教育和课程的整合这项研究计划旨在为广大学生提供更好的高性能计算教育机会，同时为计算机科学专业的学生提供并行算法设计和编程方面的即时培训。