NER: Computer Aided Design of Silicon-based Quantum Computers

华盛顿大学的马克·奥斯金量子计算机似乎是科幻小说的主题，但它们巨大的计算潜力比我们想象的更接近。尽管有很大的实际困难，但实验室已经建造了5到7位的小型量子设备。硅技术承诺更大的可扩展性。为了有效地使用这些技术，并帮助指导量子设备的研究，计算机架构师现在需要开始设计和论证量子处理器。然而，有两个主要障碍阻碍了我们。首先，基于硅的量子计算技术的简洁可描述规则尚不为人所知。其次，不存在设计、测试和评估建筑备选方案的基础设施。这项拨款涉及这两个项目。首先，我们将正式确定一种广泛讨论的、可能的量子器件技术的设计规则。其次，我们已经开始为这项技术开发体系结构抽象，可以组合在一起形成通用的量子信息处理器。第三，利用这些抽象，我们将开发一个自动化的计算机辅助设计工具，该工具将对量子计算体系结构进行描述，并将其编译成量子细胞的设备库。最后，我们将开发一个模拟器来快速评估量子系统的性价比。这些研究目标一旦实现，将成为发展量子计算机体系结构的基础资源。此外，通过对体系结构的推理，我们可以开始发现当前量子设备提议的局限性。这些限制可以转化为研究目标，也就是进一步的量子设备研究需要克服的项目。

Mark OskinUniversity of Washington0210373Quantum computers seem the subject of science fiction, but their tremendouscomputational potential is closer than we may think. Despite significantpractical difficulties, small quantum devices of 5 to 7 bits have been builtin the laboratory. Silicon technologies promise even greater scalability.To use these technologies effectively, and help guide quantum deviceresearch, computer architects need to start designing and reasoning aboutquantum processors now. However, two major hurdles stand in the way.First, compactly describable rules that characterize silicon-based quantumcomputing technologies are not known. Second, there does not exist aninfrastructure to design, test, and evaluate architectural alternatives.This grant addresses both of theseitems. First, we will formalize the design rules of a widely discussedand likely quantum device technology. Second, we have already begun todevelop architectural abstractions for this technology that can be composedtogether to form general-purpose quantum information processors. Third,using these abstractions we will develop an automated computer aideddesign tool that takes a description of a quantum computing architecture andcompiles it to a device library of quantum cells. Finally, we willdevelop a simulator that quickly estimates the cost-performance of quantumsystems.These research goals, once accomplished, will serve as foundationalresources for the development of quantum computer architectures. Inaddition, by reasoning about architectures we can begin to find the limitsof current quantum device proposals. These limits can be turned intoresearch goals, items that further quantum device research needs toovercome.