"A Quantum Etch-a-sketch": Reconfigurable quantum circuitry using chiral light

“量子蚀刻草图”：使用手性光的可重构量子电路

• 批准号: 2608166
• 金额: --
• 依托单位: University of Glasgow
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2021
• 资助国家: 英国
• 起止时间: 2021 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-2608166
• 关键词: Quantum; Etch; sketch; Reconfigurable; quantum

Superconducting wires coupled together to form Josephson junctions are used to construct quantum bits (qubits) used in quantum computers designed and built by Intel, Google, IBM and others. As with classical computers the architectures of these devices are physically fixed. Consequently, they cannot physically display neural plasticity, a prerequisite for learning, in which neural networks in brains change through growth and re-organisation. Thus, artificial neural networks are mostly run as classical algorithms. The desire to advance Artificial Intelligence (AI) has driven interest in physical neural networks, although current technologies cannot replicate physically the evolving nature of the brain. We propose a new paradigm for creating reconfigurable Josephson based qubits that can mimic the plasticity of neural networks and synaptic connections in biological systems. The concept is based on the ability to reversibly spatially control the TC of a ceramic through the exchange of angular momentum (either spin or orbital) from a light beam. The strain induced by the torque created by the exchange of angular momentum, is equivalent to that required to increase the TC of a ceramic such as yttrium barium cuprates (YBCO) by several K. Thus circuits / qubits can be reversibly created by illuminating a pattern on a macroscopic slab of ceramic, the unilluminated areas remain insulating. By changing the pattern of illumination quantum circuitry can be reconfigured. The concept presages the next step in AI offering the opportunity of self-evolving quantum computers

超导导线耦合在一起形成约瑟夫森结，用于构建量子比特(Qubit)，用于由英特尔、谷歌、IBM和其他公司设计和制造的量子计算机。与传统计算机一样，这些设备的体系结构在物理上是固定的。因此，他们无法在身体上表现出神经可塑性，这是学习的先决条件，在这种可塑性中，大脑中的神经网络通过生长和重组而改变。因此，人工神经网络大多作为经典算法运行。推动人工智能(AI)发展的愿望推动了人们对物理神经网络的兴趣，尽管目前的技术无法在物理上复制大脑的进化本质。我们提出了一个新的范例来创建可重构的基于约瑟夫森的量子比特，该量子比特可以模拟生物系统中神经网络和突触连接的可塑性。这一概念是基于通过光束的角动量(自旋或轨道)交换来可逆地控制陶瓷的TC的能力。由角动量交换产生的扭矩所产生的应变，相当于将铜酸钇(YBCO)等陶瓷的TC提高几K所需的应变。因此，可以通过照射宏观陶瓷平板上的图案来可逆地产生电路/量子比特，未被照明的区域保持绝缘。通过改变照明模式，量子电路可以重新配置。这一概念预示着人工智能的下一步，提供了自我进化的量子计算机的机会