首页> 外文学位 >Performance comparison between copper, carbon nanotube, and optics for off-chip and on-chip interconnects.
【24h】

Performance comparison between copper, carbon nanotube, and optics for off-chip and on-chip interconnects.

机译:片外和片上互连的铜,碳纳米管和光学器件之间的性能比较。

获取原文
获取原文并翻译 | 示例

摘要

For more than 30 years, the performance of silicon integrated circuits has improved at an astonishing rate. The number of functions per chip has grown exponentially, dramatically bringing down the cost per function. However, the relentless scaling paradigm is threatened by fundamental communication limits including excessive power dissipation, insufficient communication bandwidth, and large signal latency. Many of these obstacles stem from the physical limitation of Cu-based electrical wires, making it imperative to examine alternate interconnect schemes for future ICs. The two most important novel potential candidates are optical and carbon nanotube (CNT)-based interconnects.; In this dissertation, for off-chip application, we compare high speed optical and electrical interconnects using relevant metrics, such as power and bandwidth. We find that for a given communication bandwidth, beyond a critical length, power optimized optical interconnect dissipates lower power compared to high-speed electrical links. Beyond the 32nm technology node (with appropriate bandwidth requirement), optical interconnect becomes favorable for distances as little as 10cm. These distances correspond to inter-chip communication. As part of the comparisons, we also compare two different optical transmitter options comprising of a VCSEL and a modulator scheme. For the modulator option, we further presented an optimization methodology for minimizing total optical link power, and obtained the optimum quantum well modulator design parameters - number of quantum wells, pre-bias voltage, and operational parameters (swing voltage) which accomplish the low power design.; For on-chip application, we compare CNT and optical interconnects with Cu interconnects using latency, power, and a novel compound metric-bandwidth density per latency per power, which captures system requirements more efficiently. In the future, because of multi-core architecture, the designers care about the bandwidth density, latency, and power required for global communication. We extensively examine the impact of device parameters - modulator and detector capacitances for optics, materials parameters - mean free path and packing density for CNTs, and system parameters - global clock frequency and switching activity on all the aforementioned metrics.
机译:30多年来,硅集成电路的性能以惊人的速度提高。每个芯片的功能数量成倍增长,大大降低了每个功能的成本。但是,无休止的缩放范例受到基本通信限制的威胁,这些通信限制包括功耗过大,通信带宽不足以及信号延迟大。其中许多障碍源于铜基电线的物理局限性,因此有必要研究未来集成电路的替代互连方案。两个最重要的新颖潜在候选者是基于光学和碳纳米管(CNT)的互连。本文针对片外应用,利用功率和带宽等相关指标对高速光互连和电互连进行了比较。我们发现,对于给定的通信带宽,超过临界长度,与高速电气链路相比,功率优化的光学互连耗散的功率更低。除了32纳米技术节点(具有适当的带宽要求)以外,光互连对于10厘米以下的距离也变得有利。这些距离对应于芯片间通信。作为比较的一部分,我们还比较了包括VCSEL和调制器方案的两个不同的光发送器选项。对于调制器选项,我们进一步提出了一种用于使总光链路功率最小化的优化方法,并获得了最佳量子阱调制器设计参数-量子阱数量,预偏置电压和实现低功率的工作参数(摆动电压)设计。;对于片上应用,我们使用延迟,功耗以及每功耗每延迟的新型复合度量带宽密度,将CNT和光互连与Cu互连进行比较,从而更有效地捕获系统需求。将来,由于采用多核体系结构,设计人员会关心全球通信所需的带宽密度,延迟和功耗。我们广泛检查了设备参数的影响-光学器件的调制器和检测器电容,材料参数-CNT的平均自由程和堆积密度,以及系统参数-全局时钟频率和所有上述度量的开关活动。

著录项

  • 作者

    Cho, Hoyeol.;

  • 作者单位

    Stanford University.;

  • 授予单位 Stanford University.;
  • 学科 Engineering Electronics and Electrical.
  • 学位 Ph.D.
  • 年度 2007
  • 页码 175 p.
  • 总页数 175
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类 无线电电子学、电信技术;
  • 关键词

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