CIF: Small: Self-Synthesizing Mixed-signal Circuits

CIF：小型：自合成混合信号电路

• 批准号: 1319592
• 负责人: Michael Flynn
• 金额: $ 35万
• 依托单位: Regents of the University of Michigan - Ann Arbor
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2013
• 资助国家: 美国
• 起止时间: 2013-09-01 至 2017-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1319592&HistoricalAwards=false
• 关键词: CIF; Small; Self; Synthesizing; Mixed

This research addresses the bottleneck caused by the difficulty and expense of designing mixed-signal interface circuits for modern integrated circuits. These interface circuits translate between real-world analog signals and the corresponding digital signals that are processed by digital electronics. A huge challenge is that the cost of designing these mixed-signal interface circuits is skyrocketing and now dominates the overall cost of integrated circuit design. Difficulties in modeling and the worsening manufacturing variability, related to shrinking transistor size, mean that mixed-signal design requires a huge effort. Furthermore, to ensure good manufacturing yield, even the best designs are not optimal in terms of energy efficiency or performance. A fresh new approach is needed. This research investigates a completely new way of creating mixed-signal circuits. In the new approach, the integrated circuit, itself, assembles a mixed-signal function by combining and configuring simple building blocks. An integrated circuit creates a function from an uncommitted set of simple primitives and cells, such as charge-processing cells, comparators and simple amplifiers. This approach pushes design decisions and tradeoffs to each individual integrated circuit. It delivers the best possible solution for a given set of design objectives in the presence of manufacturing variations. It also avoids the need for good models and accurate simulation during design.The broader impact of this research is that it addresses the crisis of cost and complexity in mixed-signal integrated-circuit design with a radically new self-design approach. It moves design decisions to the individual integrated circuit, greatly simplifying the design process and removing mixed-signal design from the straightjacket of worst-case design. The broader impact of this research is enhanced with substantial efforts in outreach and education. These include outreach to K-12 students and research experience for undergrads from under-represented minorities. An advanced graduate-level course in analog-digital interfaces will involve graduate students in the research of mixed-signal self-design. To help change the mindset of the next generation of circuit designers, the undergrad major design experience course in analog circuits will introduce students to analog circuit optimization.

这项研究探讨了为现代集成电路设计混合信号界面电路的困难和费用所引起的瓶颈。这些界面电路在现实世界中的模拟信号和数字电子处理的相应数字信号之间转换。一个巨大的挑战是，设计这些混合信号界面电路的成本正在飙升，现在占据了集成电路设计的整体成本。与晶体管尺寸缩小有关的建模和制造变异性的困难，意味着混合信号设计需要巨大的努力。此外，为了确保良好的制造产量，即使最好的设计在能效或性能方面也不是最佳的。需要一种新的新方法。这项研究调查了一种创建混合信号电路的全新方法。在新方法中，集成电路本身通过结合和配置简单的构建块来组装混合信号功能。集成电路从一组简单的原语和单元格（例如电荷处理单元格，比较器和简单的放大器）中创建一个函数。这种方法将设计决策和权衡推向了每个单独的集成电路。它为在制造变化的存在下为给定的一组设计目标提供了最好的解决方案。它还避免了在设计过程中需要进行良好模型和准确的模拟。这项研究的更广泛影响是，它解决了具有根本新的自设计方法的混合信号集成电路设计的成本和复杂性危机。它将设计决策移至单个集成电路，从而大大简化了设计过程，并从最差的案例设计的直夹克中删除了混合信号设计。这项研究的广泛影响通过在外展和教育方面进行了大量努力增强。其中包括向K-12学生的宣传以及代表不足少数群体的本科生的研究经验。模拟数字界面的高级研究生课程将涉及研究生参与混合信号自我设计的研究。为了帮助改变下一代电路设计师的心态，模拟电路的本科生设计经验课程将向学生介绍模拟电路优化。