SBIR Phase I: Efficient Software Implementation for New Peer-to-Peer (NPP) Communications Processor

SBIR 第一阶段：新型点对点 (NPP) 通信处理器的高效软件实施

• 批准号: 0232073
• 负责人: Chitoor Srinivasan
• 金额: $ 10万
• 依托单位: EDSS., Inc.
• 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: 2003-01-01 至 2003-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0232073&HistoricalAwards=false
• 关键词: SBIR; Phase; Efficient; Software; Implementation

This Small Business Innovation Research (SBIR) Phase I project, Efficient Software Implementation of NPP Communications Processor, tests and analyses a prototype of a software implementation of a New Peer-to-Peer (NPP) asynchronous communication processor, which performs communications coordination in parallel with computations.Three areas of commercial applications for NPP are; Massively parallel High Performance Cluster Computing for real-time and non-real-time computations, Self-scheduling Parallel Programming with data distribution based on data availability, and Network-Oriented applications in finance, health, education, business and manufacturing. Efficient software implementation of NPP communications processor can dramatically decrease asynchronous communication latencies by several orders of magnitude, provide tools to dynamically debug and schedule portable parallel software systems, and provide infrastructure to produce robust software systems that can evolve dynamically. The size of this market opportunity is at least 1.2 trillion dollars over the next 10 years. NPP communications processor deployed through an API will dramatically impact this market.

这个小企业创新研究（SBIR）第一阶段项目，NPP通信处理器的高效软件实现，测试和分析了新型点对点（NPP）异步通信处理器的软件实现原型，该处理器在并行计算的同时执行通信协调。核动力发电厂的商业应用有三个领域：用于实时和非实时计算的大规模并行高性能集群计算，基于数据可用性的数据分布的自调度并行编程，以及金融、卫生、教育、商业和制造业中面向网络的应用。NPP通信处理器的高效软件实现可以将异步通信延迟显著降低几个数量级，提供动态调试和调度可移植并行软件系统的工具，并提供基础设施，以产生可动态发展的健壮软件系统。未来10年，这个市场机会的规模至少为1.2万亿美元。通过API部署的NPP通信处理器将极大地影响这个市场。