SHF: Small: Commodity Processors with Mainframe Reliability

SHF：小型：具有大型机可靠性的商品处理器

• 批准号: 1115367
• 负责人: Daniel Sorin
• 金额: $ 42万
• 依托单位: Duke University
• 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=1115367&HistoricalAwards=false
• 关键词: SHF; Small; Commodity; Processors; Mainframe

The goal of this project is to create multicore computer processorswith near-mainframe reliability at commodity costs, where the primarycosts are additional hardware and additional power consumption. Shouldthis project achieve these goals, it would transform computing.Reliable processors would no longer be a niche product; instead, theywould be the commodity products used in desktops, laptops,smartphones, etc.The research thrusts in this project include error detection, errorrecovery, diagnosis of permanent faults, and self-repair fortolerating permanent faults. The end result of this project will be ahardware prototype of a low-cost, reliable multicore processor.Building an artifact is a vital part of this project for tworeasons. First, a primary project goal is to keep power and area costslow, and these costs cannot be accurately determined using high-levelsimulation. Second, to transfer this technology to industry andtransform the industry, the chip makers must be convinced that thesolutions are viable. Chip makers have historically beenhesitant to accept ideas that have not been built, not least becauseimportant issues can be hidden in a high-level simulator.

该项目的目标是以商品成本创建具有接近大型机可靠性的多核计算机处理器，其中主要成本是额外的硬件和额外的功耗。如果这个项目实现这些目标，它将改变计算。可靠的处理器将不再是一个利基产品；相反，它们将是台式机、笔记本电脑、智能手机等中使用的商品。该项目的研究重点包括错误检测、错误恢复、永久性故障诊断以及容忍永久性故障的自我修复。 该项目的最终结果将是低成本、可靠的多核处理器的硬件原型。出于两个原因，构建工件是该项目的重要组成部分。首先，项目的主要目标是保持较低的电力和面积成本，而这些成本无法使用高级模拟来准确确定。其次，为了将这项技术转移到行业并改变行业，芯片制造商必须相信这些解决方案是可行的。 芯片制造商历来对接受尚未实现的想法犹豫不决，尤其是因为重要的问题可能隐藏在高级模拟器中。