首页> 外文期刊>IEEE transactions on circuits and systems . I , Regular papers >Circuit Designs of High-Performance and Low-Power RRAM-Based Multiplexers Based on 4T(ransistor)1R(RAM) Programming Structure
【24h】

Circuit Designs of High-Performance and Low-Power RRAM-Based Multiplexers Based on 4T(ransistor)1R(RAM) Programming Structure

机译:基于4T(ransistor)1R(RAM)编程结构的高性能,低功耗,基于RRAM的多路复用器的电路设计

获取原文
获取原文并翻译 | 示例

摘要

Routing multiplexers based on pass-transistors or transmission gates are an essential components in many digital integrated circuits. However, whatever structure is employed, CMOS multiplexers have two major limitations: 1) their delay is linearly related to the input size; 2) their performance degrades seriously when operated in near-Vt regime. Resistive Random Access Memory (RRAM) technology brings opportunities of overcoming these limitations by exploiting the properties of RRAMs and associated programming structures. In this paper, we propose new one-level, two-level and tree-like multiplexers circuit designs using 4T(ransistors)1R(RAM) elements and we compare them to naive one-level multiplexers. We consider the main physical design aspects associated with 4T1R-based multiplexers, such as the layout implications using a 7 nm FinFET technology, and the co-integration of low-voltage nominal power supply and high-voltage programming supply. Electrical simulations show that using a 7 nm FinFET transistor technology, the proposed 4T1R-based multiplexers reduce delay by 2x and energy by 2.8x over naive 4T1R and 2T1R counterparts. At nominal working voltage, considering an input size ranging from 2 to 50, the proposed 4T1R-based multiplexers reduces Area-Delay and Power-Delay products by 2.6x and 3.8x respectively, as compared to best CMOS multiplexers. In the nearVt regime, the proposed 4T1R-based multiplexer demonstrates 2x larger delay efficiency over the best CMOS multiplexer. The proposed 4T1R-based multiplexers operating at near-Vt regime can still achieve up to 22% delay improvement when compared to best CMOS multiplexers working at nominal voltage.
机译:在许多数字集成电路中,基于传输晶体管或传输门的路由多路复用器是必不可少的组件。但是,无论采用哪种结构,CMOS多路复用器都有两个主要局限性:1)它们的延迟与输入大小线性相关; 2)在接近Vt的条件下运行时,它们的性能会严重下降。电阻式随机存取存储器(RRAM)技术通过利用RRAM的特性和相关的编程结构,带来了克服这些限制的机会。在本文中,我们提出了使用4T(ransistors)1R(RAM)元件的新的一级,二级和树状多路复用器电路设计,并将它们与朴素的一级多路复用器进行了比较。我们考虑了与基于4T1R的多路复用器相关的主要物理设计方面,例如使用7 nm FinFET技术的布局含义,以及低压标称电源和高压编程电源的集成。电气仿真表明,与单纯的4T1R和2T1R同类产品相比,基于4T1R的多路复用器采用7 nm FinFET晶体管技术,可将延迟降低2倍,将能量降低2.8倍。在额定工作电压下,考虑到2至50的输入大小,与最佳CMOS多路复用器相比,基于4T1R的多路复用器建议分别将Area-Delay和Power-Delay产品减少2.6倍和3.8倍。在NearVt方案中,建议的基于4T1R的多路复用器展示出的延迟效率是最佳CMOS多路复用器的2倍。与在标称电压下工作的最佳CMOS多路复用器相比,在接近Vt体制下运行的基于4T1R的多路复用器仍可实现高达22%的延迟改善。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号