CAREER: Behavioral Modeling and Simulation of Mixed-Signal/Mixed-Technology VLSI Systems: An Integrated Research and Education Program

职业：混合信号/混合技术 VLSI 系统的行为建模和仿真：综合研究和教育项目

• 批准号: 9985507
• 负责人: C.-J. Richard Shi
• 金额: $ 26.2万
• 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: 2000
• 资助国家: 美国
• 起止时间: 2000-04-01 至 2004-03-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9985507&HistoricalAwards=false
• 关键词: CAREER; Behavioral; Modeling; Simulation; Mixed

This research aims at developing efficient techniques for modeling and simulation of complex systems-on-a-chip with millions of transistors, which contain mixed analog and digital circuits, even mixed-technology devices (microelectromechnical, optical, microfluid, etc). The focus is to explore behavioral modeling, simple-yet-accurate mathematical modeling, of a circuit in terms of its behavior. A systematic method based on symbolic network analysis is being developed to generate behavioral models from a circuit netlist. Application of traditional symbolic network analysis to large circuits is hindered by the exponential growth of the number of product terms with the size of a circuit, a problem known as term explosion. This project addresses the term problem by exploring determinant decision diagram (DDD) technology. To generate efficient simulatable code, this project explores compilation techniques to compile behavioral models and to select models of right complexity. A new paradigm for parallel and distributed simulation is being developed to exploit the maximum currency and latency in mixed analog/digital and mixed-technology systems.

这项研究旨在开发有效的技术，对具有数百万个晶体管的复杂片上系统进行建模和仿真，其中包含混合模拟和数字电路，甚至混合技术设备（微机电、光学、微流体等）。重点是探索电路行为的行为建模，即简单而准确的数学建模。正在开发一种基于符号网络分析的系统方法，用于从电路网表生成行为模型。传统符号网络分析在大型电路中的应用受到乘积项数量随电路大小呈指数增长的阻碍，这一问题称为项爆炸。该项目通过探索行列式决策图 (DDD) 技术来解决术语问题。 为了生成有效的可模拟代码，该项目探索了编译技术来编译行为模型并选择适当复杂度的模型。正在开发一种新的并行和分布式仿真范例，以充分利用混合模拟/数字和混合技术系统中的最大流通性和延迟。