Combinatorial Topology and Concurrent Computation

组合拓扑和并发计算

• 批准号: 0830491
• 负责人: Maurice Herlihy
• 金额: $ 30万
• 依托单位: Brown University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2009
• 资助国家: 美国
• 起止时间: 2009-05-01 至 2014-04-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0830491&HistoricalAwards=false
• 关键词: Combinatorial; Topology; Concurrent; Computation

Concepts and techniques drawn from combinatorial and algebraic topology have been successful at solving a number of fundamental problems common to concurrent and distributed computing. This research focuses on two goals. First, much needs to be done to further explore and exploit the emerging connection between concurrent computing and topology. In particular, there is a need to reformulate classical notions from algebraic and combinatorial topology in terms better-suited to the demands of modeling computation. Second, there are many opportunities to apply techniques based on topology to emerging models and problems, refining and extending earlier results.In the near future, nearly all computers, from supercomputers to smoke detectors, will be shared-memory multiprocessors (often called multicores). As a result, understanding how to exploit this new-found concurrency is one of the central issues in modern Computer Science.This award is co-funded by EPSCoR

从组合和代数拓扑学中提取的概念和技术已经成功地解决了并发和分布式计算中常见的一些基本问题。这项研究集中在两个目标。首先，需要做很多工作来进一步探索和利用并发计算和拓扑之间的新兴联系。特别是，有必要从代数和组合拓扑学中重新制定经典概念，以更好地适应建模计算的需求。第二，将拓扑技术应用于新出现的模型和问题，改进和扩展早期的结果，有很多机会。在不久的将来，几乎所有的计算机，从超级计算机到烟雾探测器，都将是共享内存的多处理器（通常称为多核）。 因此，理解如何利用这种新发现的并发性是现代计算机科学的中心问题之一。