Research Initiation: VLSI Architecture Implementation Transforms

研究启动：VLSI架构实现转变

• 批准号: 8806169
• 负责人: Akhilesh Tyagi
• 金额: $ 7万
• 依托单位: University of North Carolina at Chapel Hill
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 1988
• 资助国家: 美国
• 起止时间: 1988-06-01 至 1990-11-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=8806169&HistoricalAwards=false
• 关键词: Research; Initiation; VLSI; Architecture; Implementation

In the design of a VLSI architecture, there are several choices of performance parameters for each sub-system available to a designer. The choice of design parameters is so complex that very seldom is the whole design space explored. To explore alternative designs, the only recourse a designer has is to redesign the target parts completely. Dr. Tyagi is investigating automation of this process. He is exploring a general set of design transforms that, when applied to a given design with a set of performance parameters (in terms of area, energy, and time), can produce a design with the desired performance. The principal investigator's initial research indicates that such a set of transforms can be identified and implemented. He is applying the transforms to designing functions such as shifting, multiplication and sorting, and finding techniques to compose these functions into an architecture with given performance characteristics.

在超大规模集成电路架构的设计中， 每个子系统的性能参数， 设计师 设计参数的选择非常复杂， 很少有人探索整个设计空间。 探讨 替代设计，设计师唯一的追索权是 重新设计目标部件。 泰亚吉医生正在调查 这个过程的自动化。 他正在探索一套通用的 设计转换，当应用于给定的设计时， 性能参数（在面积、能量和时间方面）， 可以产生具有所需性能的设计。 的 首席研究员的初步研究表明，这样一个 可以识别和实现一组变换。 他是 将变换应用于设计换档等功能， 乘法和排序，并找到技术组成 将这些功能整合到具有给定性能的体系结构中 特色