CAREER: Robust Design for Scaled CMOS Technologies

Prop No：0643319标题：Career：Robust Design for Scaled CMOS TechnologiesPI：Subhasish MitraInst：Stanford University摘要本项目专注于为设计稳健的计算和通信系统创造新的核心技术，并培养未来一代工程师，他们将负责以稳健性为基本设计目标来设计此类系统。这些系统构成了当今人类生活中不可分割的一部分，并将在未来进一步整合到广泛的复杂应用中，如信息家电、交通、健康、生物医疗和金融。该项目的动机是，由于未来半导体工艺技术中日益严重的硬件故障问题，硬件设计即将发生范式转变。传统的硬件设计范式假设半导体工艺技术总是产生完美的硬件，在其生命周期内永远不会错误地运行。半导体行业成功的关键是不断推动更小的半导体工艺几何尺寸，也称为技术规模，这一点正接近这样的设计范例将由于基本限制而不可行的地步。该项目通过采用以下原则打破了这一障碍：接受底层硬件将不完美的事实，并开发技术和工具来设计高效、经济高效和故障感知的健壮系统，以确保高服务质量、数据完整性、可用性、安全性和安全性。计算机工程的这一进步将使半导体行业能够继续扩大技术规模，对未来的电子产品产生巨大影响。预计该项目中形成的研究思路将对新兴革命性纳米技术中的稳健设计产生深远影响，远远超出当代的cmos技术。

Prop No: 0643319Title: CAREER:Robust Design for Scaled CMOS TechnologiesPI: Subhasish MitraInst: Stanford UniversityAbstractThis project focuses on creating novel core technologies for designing robust computing and communication systems, and educating future generations of engineers who will be responsible for designing such systems with robustness as a fundamental design objective. Such systems constitute an inseparable part of human lives today, and will be further integrated in the future over a wide range of complex applications such as information appliances, transportation, health, bio-medical, and finance. The project is motivated by an imminent paradigm shift in hardware design resulting from the growing problem of hardware failures in future semiconductor process technologies. The traditional hardware design paradigm assumes that semiconductor process technologies always produce perfect hardware that will never operate incorrectly during its lifetime. Continuous push for smaller semiconductor process geometries also called technology scaling which is the key to the success of the semiconductor industry is approaching a point where such a design paradigm will be infeasible due to fundamental limitations. This project breaks this barrier by adopting the following principle: accept the fact that the underlying hardware will be imperfect, and develop technologies and tools for designing efficient, cost-effective, and failure-aware robust systems that ensure high quality of service, data integrity, availability, security and safety. This advancement in Computer Engineering will enable the semiconductor industry to continue technology scaling with immense impact on future electronic products. It is also expected that the research ideas developed in this project will have a far-reaching impact on robust designs in emerging revolutionary nanotechnologies far beyond contemporary CMOS technologies.