SHF: Small: Device-Circuit-Architecture Co-Design for Reconfigurable Computing at the Extreme

SHF：小型：用于极限可重构计算的器件-电路-架构协同设计

• 批准号: 1116102
• 负责人: Swarup Bhunia
• 金额: $ 40万
• 依托单位: Case Western Reserve 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: 2011-09-01 至 2015-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1116102&HistoricalAwards=false
• 关键词: SHF; Small; Device; Circuit; Architecture

The objective of this project is to explore technology for low-power reconfigurable computing at extreme environment, i.e. at a temperature 250°C and at high radiation (1-30 Mrad), where conventional electronics fail to work reliably. The project seeks to develop the computing platform through a device-circuit-architecture co-design approach. The platform uses polycrystalline SiC nanoelectromechanical system (NEMS) switches to achieve nanoscale dimension, superior mechanical and chemical stability at extreme conditions and virtually zero leakage. The project studies a novel switch structure with multi-layer cantilever beams to realize the reconfigurable building blocks with high density and robustness. Possible hybridization with SiC junction field effect transistor (JFET) will be explored to address the issue of cascading mechanical logic blocks with limited driving capability. The project also investigates a novel computational model and architecture to adapt to failures and to achieve low interconnect overhead. Considering the lack of simulation models for NEMS switches and circuits, it will develop appropriate circuit-compatible models. Operation of the building blocks at high temperature will be validated through simulation as well as fabrication and measurements of test chips. The proposed low-power reconfigurable hardware design approach can provide enabling technology for extreme-environment computing in number of application areas including automotive and industrial applications, space, avionics, combustion engine, and intelligent propulsion systems. It can provide an order of magnitude lower power and area than state-of-the-art SiC transistor based electronics. The simulation models can be valuable research tool for analysis of nanoscale mechanical switches. The research will integrate education and training through development of a new undergraduate course, interdisciplinary senior projects, and an internet group on harsh environment electronics.

该项目的目标是探索在极端环境下的低功耗可重构计算技术，即在250°C的温度和高辐射(1-30mrad)下，传统电子设备无法可靠地工作。该项目寻求通过设备-电路-架构共同设计的方法来开发计算平台。该平台使用多晶碳化硅纳米机电系统(NEMS)开关来实现纳米尺寸，在极端条件下具有优异的机械和化学稳定性，并且几乎没有泄漏。该项目研究了一种新颖的多层悬臂梁开关结构，以实现高密度和健壮性的可重构积木。将探索与碳化硅结场效应晶体管(JFET)杂交的可能性，以解决驱动能力有限的级联机械逻辑块的问题。该项目还研究了一种新的计算模型和体系结构，以适应故障并实现低互连开销。考虑到目前缺乏NEMS开关和电路的仿真模型，它将开发合适的电路兼容模型。将通过模拟以及测试芯片的制造和测量来验证构建块在高温下的操作。提出的低功耗可重构硬件设计方法可以在汽车和工业应用、空间、航空电子、内燃机和智能推进系统等多个应用领域为极端环境计算提供使能技术。它可以提供比最先进的基于碳化硅晶体管的电子产品低一个数量级的功率和面积。仿真模型可以作为分析纳米级机械开关的有价值的研究工具。这项研究将通过开发一门新的本科课程、跨学科的高级项目以及一个关于恶劣环境电子的互联网小组来整合教育和培训。