NANO: Applications, Architectures, and Circuit Design for Nano-scale Magnetic Logic Devices

磁系统对计算逻辑很有吸引力，因为它们是非易失性的，并且提供了每个逻辑操作的低功耗的前景。最近的工作表明，纳米磁体可以(1)被配置成实现通用逻辑门(可以用来实现任何布尔函数)，并且(2)不具有早期笨重的铁氧体磁体的缺点。这项工作将研究量子点细胞自动机(QCA)体系结构方案中光刻定义的纳米磁体，其中单个纳米磁体之间的直接物理相互作用产生逻辑功能。本文提出的工作旨在从研究逻辑门跳到研究简单电路。反过来，分析电路应该提供对由磁性QCA(MQCA)设备制成的简单计算系统的预期性能的重要洞察。系统是理想的最终结果。该提案有三个主要目标：1.调查mQCA设备系统的潜在应用空间。为两个看似很合适的应用程序空间构建所需的核心逻辑。使用(1)和(2)产生的设计、见解和实验数据来确定mQCA器件是否能够在系统级别上超越基于半导体的同等器件。在器件密度、功耗、功率密度、对热波动的容忍度和全球带宽方面，纳米级磁体的性能似乎与(并且往往优于)路线图末端的CMOS。随着这项技术的基本组成部分已经被实验证明，确定这项技术的可行性的一些挑战现在转移到了计算机科学家身上。然而，这项工作也将讨论将涉及制造系统的“核心”部分的实验。在系统的背景下，磁性材料可以提供制造的简单性、坚固性和真正的室温操作。这可能建议为mQCA提供需要强大性能和低功耗的应用程序空间。磁性材料对辐射也不敏感，这可能意味着在恶劣的操作环境中，如外太空，或用于卫星和军事应用时，磁性材料具有优异的性能。在系统层面，这项工作将以数字信号处理和可重新编程逻辑为目标。在实验上，这项工作将研究I/O结构、高效互连和可编程多数门。最后，这项研究不仅将为磁性QCA的可行性提供重要的见解，而且许多拟议的工作也是独立于实现的-并且也将特别适用于分子QCA设备系统。

Magnetic systems are attractive for computational logic because they are nonvolatile and offer the promise of low power dissipation per logic operation. Recent work shows that nanomagnets can (1) be configured to realize a universal logic gate (which can be used to implement any Boolean function) and (2) do not possess the disadvantages of the early, bulky, ferrite core magnets. This work will investigate lithographically defined nanomagnets within the quantum-dot cellular automata (QCA) architecture scheme - where the direct physical interaction between individual nanomagnets yields logic functionality.The work proposed here aims to make the jump from studying logic gates to studying simple circuits. Analyzing circuits should in turn provide significant insight as to the expected performance of simple computational systems made from magnetic QCA (mQCA) devices. Systems are the desired end result. The proposal has three main goals:1. To investigate potential application spaces for systems of mQCA devices.2. To fabricate the core logic needed for two seemingly well-suited application spaces.3. To use the designs, insights, and experimental data produced by (1) and (2) to determine whether or not mQCA devices can outperform semiconductor-based equivalents at the systems-level.The performance of nano-scale magnets appears to be competitive with (and is often better than) end-of-the-roadmap CMOS in the context of device density, power consumption, power density, tolerance to thermal fluctuations, and global bandwidth. As the basic components of this technology have been experimentally demonstrated, some of the challenges in determining the viability of this technology now shift to computer scientists. However, this work will also discuss experiments that will involve the fabrication of the "core" parts of the systems to be explored.In the context of systems, magnetic materials could offer simplicity of fabrication, robustness, and true room temperature operation. This might suggest application spaces for mQCA that require robust performance and low power consumption. Magnetic materials are also insensitive to radiation, which might suggest superior performance in harsh operating environments such as outerspace, or for satellite and military applications. At the systems level, this work will target digital signal processing and reprogrammable logic. Experimentally, this work will investigate an I/O structure, efficient interconnect, and a programmable majority gate.Finally, not only will this study provide significant insight as to the viability of magnetic QCA, but much of the proposed work is also implementation independent - and will apply particularly well to systems of molecular QCA devices too.