TWC: Large: Collaborative: Verifiable Hardware: Chips that Prove their Own Correctness

TWC：大型：协作：可验证的硬件：证明自身正确性的芯片

• 批准号: 1801052
• 负责人: Michael Taylor
• 金额: $ 63.72万
• 依托单位: University of Washington
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2017
• 资助国家: 美国
• 起止时间: 2017-09-01 至 2023-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1801052&HistoricalAwards=false
• 关键词: TWC; Large; Collaborative; Verifiable; Hardware

This project addresses how semiconductor designers can verify the correctness of ICs that they source from possibly untrusted fabricators. Existing solutions to this problem are either based on legal and contractual obligations, or use post-fabrication IC testing, both of which are unsatisfactory or unsound. As a sound alternative, this project designs and fabricates verifiable hardware: ICs that provide proofs of their correctness for every input-output computation they perform in the field. These proofs must be efficiently verifiable in less time and energy than it takes to re-execute the computation itself.Building upon exciting recent theoretical and practical advances in verifiable outsourced computation for the cloud, this project develops new techniques that exploit the unique constraints and adversary models that relate to the verifiable hardware problem. In addition, the project also develops new practical approaches to the problem of general verifiable computation. As a broader impact, computing systems security is one of the greatest technological problems faced by society today. Verifiable hardware is an essential foundation for building future computing systems that are reliable and free from catastrophic security failures. The ultimate goal of this project is to make verifiable hardware practical and accessible for use in cryptographic and mission-critical hardware applications through open-source tools. The PIs are strongly committed to education and public outreach by producing widely-used course materials and taking active roles in outreach at minority-serving universities, community colleges, student organizations and high schools.

该项目介绍了半导体设计师如何验证他们可能是从可能不受信任的制造商那里源的IC的正确性。 现有解决此问题的解决方案是基于法律和合同义务，或者使用后制作IC测试，这两者都不令人满意或不合适。作为一种声音替代方案，该项目设计并制造了可验证的硬件：IC，可为其在现场执行的每个输入输出计算提供正确性的证明。这些证明必须在更少的时间和精力中有效地进行比较重新执行计算本身所需的证明。建立令人兴奋的云的可验证外包计算的令人兴奋的理论和实际进步，该项目开发了新技术，可以利用与可验证的硬件问题相关的独特约束和对象模型。此外，该项目还为一般可验证的计算问题开发了新的实用方法。作为更广泛的影响，计算系统安全是当今社会面临的最大技术问题之一。可验证的硬件是建立可靠且没有灾难性安全故障的未来计算系统的重要基础。该项目的最终目标是使可验证的硬件实用，并且可以通过开源工具用于加密和关键任务硬件应用程序。 PI通过生产广泛使用的课程材料并在少数派大学的大学，社区学院，学生组织和高中发挥积极作用，从而强烈致力于教育和公众推广。