Collaborative Research: Globally Optimized Robust Systems on Multi-Core Hardware

协作研究：多核硬件上的全局优化鲁棒系统

• 批准号: 0903459
• 负责人: Subhasish Mitra
• 金额: $ 19万
• 依托单位: Stanford 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: 2009-08-01 至 2012-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0903459&HistoricalAwards=false
• 关键词: Collaborative; Research; Globally; Optimized; Robust

The focus of this project is on designing robust multi-core systems at lowest cost that are resilient to hardware failures. It is motivated by an imminent paradigm shift in hardware design resulting from the growing problem of hardware failures in future process technologies. The central vision is to develop techniques and tools for designing hierarchical robust multi-core systems that are globally optimized across multiple abstraction layers ? circuit, architecture, runtime, and application ? without incurring the high cost of expensive redundancy techniques. An inter disciplinary approach will integrate research and education required to enable future robust systems. Because of its cross-cutting nature, the proposed research has the potential of having a major impact on future systems, enabling future technology scaling, thus making everyday lives better.

该项目的重点是以最低的成本设计强大的多核系统，这对硬件故障有抵抗力。这是由于硬件设计的迫在眉睫的范式转变，这是由于未来过程技术中硬件故障的日益增长的问题所致。中心愿景是开发用于设计层次稳健多核系统的技术和工具，这些系统在多个抽象层之间进行了全球优化？电路，架构，运行时和应用程序？不产生昂贵的冗余技术的高成本。纪律间的方法将整合所需的研究和教育，以实现未来的健壮系统。由于其跨裁切性质，拟议的研究可能会对未来的系统产生重大影响，从而实现未来的技术扩展，从而使日常生活更好。