BSF:2012139:Computing Structures Beyond Moore and von Neumann

该项目是美国-以色列计算机科学合作（USICCS）计划的一部分。 通过该计划，NSF和美国-以色列两国科学基金会（BSF）共同支持美国研究人员和以色列研究人员之间的合作。计算系统嵌入在各种各样的商业应用中，推动经济发展，改变我们的社会和文化。由于对可移植性和不断增加的复杂性的需求，下一代计算系统正变得受到能量耗散约束的严重限制。这些计算机系统的应用范围从电池操作的个人设备到服务于互联网的大规模数据服务器。该项目的主要目标是提供电路和架构设计技术，通过在这些计算系统的设计中利用新的磁技术，从根本上增强这种能源约束条件。具体来说，忆阻器将被用于开发先进的计算系统，这些系统将使用更少的功率，同时由于这种新兴和令人兴奋的技术的固有优势，提供前所未有的新应用。忆阻器在没有能量的情况下保持状态的能力将被用作片上存储器的替代品;例如，DRAM和高速缓存，以大大降低便携式计算的能量需求。此外，由于容易地将忆阻器与CMOS集成的能力，将开发混合电路和架构以提供与现代计算系统相比不同地且有利地操作的新颖计算结构。这些新颖的电路结构、计算架构和先进的磁技术的结合将有助于更好地理解和利用片上信息的局部性，同时重振计算技术的缓慢发展。此外，在节省能源和改善环境的同时，将实现更好的便携性。

This project is funded as part of the United States-Israel Collaboration in Computer Science (USICCS) program. Through this program, NSF and the United States - Israel Binational Science Foundation (BSF) jointly support collaborations among US-based researchers and Israel-based researchers. Computing systems are embedded in a wide variety of commercial applications, driving the economy and transforming our society and culture. The next generation computing systems are becoming severely limited by energy dissipation constraints, due to the need for portability and ever increasing complexity. These computer systems range in applications from battery operated personal devices to large scale data servers servicing the internet. The primary objectives of this project are to provide circuit and architectural design technologies to fundamentally enhance this condition of energy constraints by exploiting new magnetic technologies in the design of these computing systems. Specifically, memristors will be exploited to develop advanced computing systems that will use fundamentally less power while offering novel heretofore unseen applications due to the inherent advantages of this burgeoning and exciting technology.The ability of memristors to maintain state without energy will be exploited as a replacement for on-chip memory; for example, DRAM and high speed cache, to greatly lower the energy requirements of portable computing. Furthermore, due to the ability to easily integrate memristors with CMOS, hybrid circuits and architectures will be developed to provide novel computing structures that operate differently and advantageously as compared to modern computing systems. The combination of these novel circuit structures, computing architectures, and advanced magnetic technologies will contribute to better understanding and exploitation of on-chip locality of information while reinvigorating the slowing development of computing technologies. Furthermore, improved portability will be attained while saving energy and enhancing the environment.