SHF: Small: Reconditioning: Optimizing Conditional Communication in Asynchronous Design

SHF：小：修复：优化异步设计中的条件通信

• 批准号: 1116416
• 负责人: Peter Beerel
• 金额: $ 30万
• 依托单位: University of Southern California
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2011
• 资助国家: 美国
• 起止时间: 2011-07-01 至 2015-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1116416&HistoricalAwards=false
• 关键词: SHF; Small; Reconditioning; Optimizing; Conditional

This proposal addresses the grand challenge of reducing power consumption of digital electronic circuits leveraging off of a promising new automated design flow for asynchronous circuits. This novel flow actively manages data-flow within the circuit and is capable of disabling computational blocks that would otherwise perform unnecessary calculations and waste power. The proposed research is to develop mathematical software algorithms that will optimize the partitioning of the circuit into computational blocks to maximize the blocks that are disabled and thereby minimize power consumption.The impact of this research will span engineering, scientific, and social values. First, it will enable synthesized asynchronous designs to be lower-power and faster and thus more compelling. Moreover, because the research is based on a formal framework of communicating processes it is generally applicable to many computational systems, including next-generation beyond-CMOS computational platforms. At the social level it will enable the enrichment of students at USC and other institutions as the tool flow is made more widely available. Lastly, it will provide a rich context for a pre-college outreach program in which the principal investigator teaches high-school students about the world of electrical engineering, integrated chip design, and in a greener economy and helping form a sustainable society

该提案解决了降低数字电子电路功耗的重大挑战，利用了一种有前途的新的异步电路自动设计流程。这种新颖的流主动管理电路内的数据流，并且能够禁用计算块，否则这些计算块将执行不必要的计算并浪费功率。该研究旨在开发数学软件算法，优化电路划分为计算模块，以最大限度地提高禁用的模块，从而最大限度地降低功耗。该研究的影响将涵盖工程，科学和社会价值。首先，它将使合成异步设计更低功耗和更快，从而更引人注目。此外，由于该研究基于通信过程的正式框架，因此它通常适用于许多计算系统，包括下一代超CMOS计算平台。在社会一级，随着工具流的广泛提供，它将使南加州大学和其他机构的学生能够充实自己。最后，它将为大学预科外展计划提供丰富的背景，在该计划中，首席研究员向高中生讲授电气工程，集成芯片设计，绿色经济和帮助形成可持续发展的社会