CAREER: Provably Good Concurrency Platforms for Streaming Applications

职业：经过验证的流应用程序良好并发平台

• 批准号: 1150036
• 负责人: Kunal Agrawal
• 金额: $ 42.29万
• 依托单位: Washington University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2012
• 资助国家: 美国
• 起止时间: 2012-07-01 至 2018-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1150036&HistoricalAwards=false
• 关键词: CAREER; Provably; Good; Concurrency; Platforms

Most of the today's computers are parallel machines. For example, multicore processorson personal computers and cell phones, graphics processors, cloudcomputers, and clusters all have more than one processing unit. Many modern applications are data intensive; examples include digital signal processing, audio and videoprocessing, networking, scientific and biological computations.Streaming is an increasingly popular paradigm for applications that process large amounts of data in parallel. A streaming concurrency platform is responsible forcorrectly and efficiently executing streaming applications on a given parallel machine.The goal of this research is to design streaming concurrency platformsthat provide guarantees of forward progress and efficiency. Thisresearch will fundamentally advance the technology by addressing the following important questions: (1) How to guarantee that applications will make forward progress (notdeadlock)? (2) How to guarantee that the streaming applications will run efficientlyon modern parallel machines with deep and complex memory hierarchies?and(3) How to support more general and expressive streaming models whilestill providing correctness and performance guarantees? In modern machines cache locality can have a significant impact onperformance. This research has the potential to make fundamentalcontributions to both design and analysis techniques for steamingschedulers that guarantee good cache performance to streaming applications.This work will enable programmers to express a larger class of applications moreeasily in the streaming model. While this research project is a primarily theoretical undertakingaimed towards designing algorithms and proving asymptotic bounds ontheir performance, the overall goal is to enable practical andefficient concurrency platforms that can run high-performance,high-throughput streaming computations on real parallel machines.Therefore, one of the primary objectives will be to design low overheadand simple algorithms that can be implemented in production-levelstreaming concurrency platforms.

当今的大多数计算机都是并行计算机。 例如，个人电脑和手机上的多核处理器、图形处理器、云计算机和集群都具有多个处理单元。 许多现代应用程序都是数据密集型的；示例包括数字信号处理、音频和视频处理、网络、科学和生物计算。对于并行处理大量数据的应用程序来说，流式传输是一种日益流行的范例。 流式并发平台负责在给定的并行机器上正确、高效地执行流式应用程序。本研究的目标是设计能够提供前进进度和效率保证的流式并发平台。 这项研究将从根本上推进技术的发展，解决以下重要问题：（1）如何保证应用程序向前进展（而不是死锁）？ （2）如何保证流应用程序能够在具有深度和复​​杂的内存层次结构的现代并行机器上高效运行？（3）如何支持更通用和更具表现力的流模型，同时仍然提供正确性和性能保证？在现代机器中，缓存局部性会对性能产生重大影响。 这项研究有可能对流调度器的设计和分析技术做出根本性贡献，从而保证流应用程序具有良好的缓存性能。这项工作将使程序员能够在流模型中更轻松地表达更大类的应用程序。 虽然该研究项目主要是一项理论工作，旨在设计算法并证明其性能的渐近界限，但总体目标是实现实用且高效的并发平台，可以在真正的并行机上运行高性能、高吞吐量的流计算。因此，其中之一主要目标是设计可在生产级流并发平台中实现的低开销且简单的算法。