Energy-Efficient and Reconfigurable Network-on-Chips for IP Integration in Complex SoC Systems

0702168杨梅，内华达大学拉斯维加斯分校用于复杂SoC系统中IP集成的节能可重构片上网络智能优势：本项目专注于研究节能可重构片上网络(NoCs)，以解决复杂片上系统(SoC)系统中知识产权(IP)集成所面临的主要挑战，包括能效、可重构性、可扩展性和信号完整性。拟议工作的目标包括：研究基于可重构递归对角环(RDT)的NoC体系结构；研究了多路径路由方案；研究优先的虫洞切换技术；以及。开发原型系统以测试所有提出的解决方案，并在构建原型时解决系统集成问题。本文的研究成果包括：(1)对可重构片上网络体系结构的研究有助于理解片上网络的一些基本问题；(2)提出的路由方案是在网络级解决信号完整性问题的首次尝试，这将推动片上网络跨层优化通信协议的研究；(3)通过原型系统的设计，研究片上网络中具有挑战性的实现问题。更广泛的影响：拟议研究的更广泛影响包括：(1)拟议的工作将为设计(在架构选择和通信模型方面)和综合未来的计算密集型SoC系统开辟一条新的途径；(2)该项目将通过项目参与和课程开发，提高联合国志愿人员组织欧洲经委会系研究生和本科教育的质量；(3)社区外联方案将对高中生，特别是那些来自少数群体和代表性不足的学生，在科学和工程领域选择有价值的职业产生深远的影响；(4)联合国志愿人员方案正在开发R&AMP；SoC新兴领域的研发能力将帮助建立基本的科学和技术合作伙伴关系，使技术从大学研究快速转移到行业，并吸引和留住有才华的学生在该州学习和就业，从而显著升级和多样化内华达州的经济。

0702168Mei Yang, University of Nevada-Las VegasEnergy-Efficient and Reconfigurable Network-on-Chips for IP Integrationin Complex SoC SystemsIntellectual Merit: This project focuses on investigating energy-efficient and reconfigurable Network-on-Chips (NoCs) to address the major challenges faced by Intellectual Property (IP) integration in complex System-on-Chip (SoC) systems, including energy efficiency, reconfigurability, scalability, and signal integrity. The objectives of the proposed work include: . Investigate reconfigurable Recursive Diagonal Torus (RDT)-based NoC architectures;. Investigate multi-path routing scheme;. Investigate the prioritized wormhole switching technique; and . Develop prototype systems to test all the proposed solutions and tackle the system integration issues in building the prototypes. The intellectual merit of the proposed research includes: (1) The study of reconfigurable NoC architectures will help to understand some fundamental issues related to the NoC paradigm; (2) The proposed routing scheme is the first attempt in addressing signal integrity problem at the network level, which will promote the research in developing cross-layer optimized communication protocols for NoCs; and (3) Challenging implementation issues related to NoCs are studied by the designs of the prototype systems. Broader Impacts: The broader impacts of the proposed research include: (1) The proposed work will open a new avenue for designing (in architecture selection and communication model) and synthesizing future computation-intensive SoC systems; (2) The project will improve the quality of the graduate and undergraduate education in the ECE department at UNLV through project participation and curriculum development; (3) The community outreach program shall have profound impacts on high school students, especially those students from the minority and the underrepresented groups, to choose a rewarding career in science and engineering; and (4) UNLV's developing R&D capabilities in the emerging area of SoCs will significantly escalate and diversify Nevada economy by helping establish essential science and technology partnerships, allowing fast technology transfer from the university research to industry, and attracting and keeping the talented students to study and be employed in the state.