SHF: Small: Methodology, Tools, and Circuits for Bundled-Data Resilient Asynchronous Design

SHF：小型：用于捆绑数据弹性异步设计的方法、工具和电路

• 批准号: 1619415
• 负责人: Peter Beerel
• 金额: $ 40万
• 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: 2016
• 资助国家: 美国
• 起止时间: 2016-07-01 至 2020-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1619415&HistoricalAwards=false
• 关键词: SHF; Small; Methodology; Tools; Circuits

The impact of this research will span both academia and industry, yielding: (1) fundamental theory to design and analyze asynchronous circuits that gracefully adapt to process variability and makes near-threshold (i.e., low-power) computing practical; (2) a CAD flow that makes asynchronous circuits far more attractive for the ultra-low-power market. The research is coupled with a comprehensive plan to foster innovation in engineering, including plans for commercializing this research. The plans also include student training and future workforce development for the electronic chip industry.The unrelenting demand for longer battery life and energy-efficient designs is increasing the desire for integrated chip voltage supplies to go lower and lower, approaching near threshold levels. This aggravates process variability which forces increasing margins in traditional synchronous clock periods, yielding far-from-optimal solutions. The focus of this proposal is a recently proposed new resilient asynchronous design style, called the "Blade", that is robust to meta-stability, robust to hold times, and architecturally agnostic, not relying on architectural replay to correct for errors. This proposal will create a comprehensive framework to support the efficient design of Blade circuits with emphasis on test, synthesis, and physical design.

这项研究的影响将涵盖学术界和行业，产生：（1）设计和分析异步电路的基本理论，这些电路优雅地适应了过程可变性，并使接近阈值（即低功率）计算实用； （2）使异步电路的CAD流对超低功率市场更具吸引力。这项研究与促进工程创新的全面计划相结合，包括将这项研究商业化的计划。该计划还包括针对电子芯片行业的学生培训和未来的劳动力发展。对更长的电池寿命和节能设计的不懈需求正在增加对综合芯片电压供应的渴望，以越来越降低，接近阈值水平。这种加剧了过程变异性，这会迫使传统同步时钟周期中增加边缘，从而产生了远距离最佳的解决方案。 该提案的重点是最近提出的一种新的弹性异步设计样式，称为“刀片”，对元稳定性非常强大，可以保持时间稳健，并且在建筑上不可知，而不是依靠建筑重播来纠正错误。该建议将创建一个全面的框架，以支持刀片电路的有效设计，重点是测试，合成和物理设计。