Methods towards photonic quantum error detection and correction

光子量子错误检测和校正的方法

• 批准号: MR/W011794/1
• 负责人: Raj Patel
• 金额: $ 181.35万
• 依托单位: Imperial College London
• 依托单位国家: 英国
• 项目类别: Fellowship
• 财政年份: 2022
• 资助国家: 英国
• 起止时间: 2022 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=MR%2FW011794%2F1
• 关键词: Methods; towards; photonic; quantum; error

What would computer pioneers Mauchly and Eckert's reaction be if they could take a glimpse at life in 2021? In 1945 they built ENIAC, the first general-purpose digital computer-which took up an entire room. That processing power is now completely eclipsed by devices that can fit on our wrist! Despite tremendous progress over the past 50 years, a new paradigm is on the horizon which will shatter the way we think about information and deliver new computing technologies limited only by nature herself. Laptops, tablets, and phones-devices we rely on every day-consist of billions of logic gates forming the building blocks of processors. They process binary data (0s and 1s) in a meaningful way. While serving us very well so far, certain tasks easily push current classical technology to its absolute limit. Examples include solving hard equations, big-data analysis, simulating complex molecules for drug discovery and predicting real-time changes in weather and financial markets. These all require serious computational grunt and in some cases are impossible to complete even with the fastest classical supercomputers but are made possible with quantum computers. The first quantum revolution occurred when physicists discovered that the laws of nature affect atomic-scale objects differently compared to everyday objects we interact with. This led scientists to propose the coming of a second quantum revolution, one which harnesses the properties of electrons, atoms, and photons-single particles of light-to create powerful new technologies. The allure of quantum computers is the unparalleled processing power that they could provide. They use quirky effects such as superposition, where a particle can be in two places at once, and entanglement where observing one particle can provide information about another, no matter how far apart they are. We can see these counterintuitive but very real effects in the properties of photons such as their position, time of arrival, frequency, or polarisation and define special quantum bits or qubits. With their ability to travel fast and far, and well-established techniques to create, manipulate and detect them, photons make a favourable choice for qubits. Qubits are however much more fragile than their classical counterparts and quantum information can easily be corrupted by loss or by improper preparation or manipulation of photons. Qubits are manipulated by circuits-much like our everyday computer. These errors can quickly stack up for large numbers of qubits and operations which are often required for tackling difficult computational tasks. I will investigate strategies to detect and correct errors in photonic quantum processors. Much like electrical circuits in classical computers, I will use chips containing circuits, only these circuits carry optical signals in the form of qubits encoded onto pulses of light. I will use chips designed to limit loss together with a new novel device designed to improve the collection of the light. Errors arising from faults in the circuitry can be detected using multiple copies of the circuit with the effect of routing these errors to specific parts of the chip where they can be discarded. Additionally, I will engineer exotic states of light called 'GKP' states (after inventors Gottesman, Kitaev and Preskill) which can inherently cope with a broad class of errors and serve as a key building block towards creating scalable universal fault-tolerant quantum computers. While GKP states have been known for almost 20 years, limitations in photonic hardware have prohibited their creation in the lab. However, recent innovations in light sources, circuits, and detectors by my group together with new theoretical insight by physicists around the world bring us within grasp of these important quantum states of light. These routes to handling errors will be vital to realising powerful photonic quantum technologies and will spur further innovations in the near future.

如果计算机先驱莫奇利和埃克特能看到2021年的生活，他们会有什么反应？1945年，他们建造了第一台通用数字计算机ENIAC，占据了整个房间。这种处理能力现在完全被可以戴在手腕上的设备所掩盖！尽管在过去的50年里取得了巨大的进步，但一个新的范式即将出现，它将打破我们对信息的思考方式，并提供仅受自然本身限制的新计算技术。笔记本电脑、平板电脑和手机——这些我们每天都依赖的设备——由数十亿个逻辑门组成，这些逻辑门构成了处理器的构建模块。它们以有意义的方式处理二进制数据（0和1）。虽然到目前为止为我们提供了很好的服务，但某些任务很容易将当前的经典技术推向其绝对极限。例子包括解决困难的方程、大数据分析、模拟用于药物发现的复杂分子，以及预测天气和金融市场的实时变化。这些都需要大量的计算能力，在某些情况下，即使是最快的经典超级计算机也无法完成，但量子计算机使之成为可能。第一次量子革命发生时，物理学家发现自然定律对原子尺度物体的影响与我们日常接触的物体不同。这使得科学家们提出了第二次量子革命的到来，即利用电子、原子和光子（光的单个粒子）的特性来创造强大的新技术。量子计算机的魅力在于它可以提供无与伦比的处理能力。他们使用了一些奇怪的效应，比如叠加，一个粒子可以同时出现在两个地方，以及纠缠，观察一个粒子可以提供关于另一个粒子的信息，不管它们相距多远。我们可以在光子的属性中看到这些违反直觉但非常真实的影响，比如它们的位置、到达时间、频率或极化，并定义特殊的量子比特或量子位。光子有能力传播得又快又远，并且有成熟的技术来制造、操纵和探测它们，因此光子是量子比特的一个有利选择。然而，量子比特比它们的经典对偶要脆弱得多，量子信息很容易因光子的丢失或不适当的制备或操作而损坏。量子比特是由电路操纵的，就像我们日常使用的计算机一样。在处理困难的计算任务时，通常需要大量的量子比特和操作，这些错误会迅速累积起来。我将研究在光子量子处理器中检测和纠正错误的策略。就像经典计算机中的电路一样，我将使用包含电路的芯片，只有这些电路以编码于光脉冲上的量子比特的形式携带光信号。我将使用设计用于限制损耗的芯片以及设计用于改善光收集的新型设备。由电路故障引起的错误可以使用电路的多个副本来检测，其效果是将这些错误路由到芯片的特定部分，在那里它们可以被丢弃。此外，我将设计被称为“GKP”状态的奇异光态（以发明家Gottesman， Kitaev和Preskill命名），它可以固有地处理各种各样的错误，并作为创建可扩展的通用容错量子计算机的关键构建块。虽然GKP态已经被发现近20年了，但光子硬件的限制阻碍了它们在实验室中的创造。然而，我的团队最近在光源、电路和探测器方面的创新，以及世界各地物理学家的新理论见解，使我们能够掌握这些重要的光的量子态。这些处理错误的途径对于实现强大的光子量子技术至关重要，并将在不久的将来刺激进一步的创新。

项目成果

A universal programmable Gaussian boson sampler for drug discovery.

• DOI: 10.1038/s43588-023-00526-y
• 发表时间: 2023-10
• 期刊: Nature computational science
• 作者: Yu S;Zhong ZP;Fang Y;Patel RB;Li QP;Liu W;Li Z;Xu L;Sagona-Stophel S;Mer E;Thomas SE;Meng Y;Li ZP;Yang YZ;Wang ZA;Guo NJ;Zhang WH;Tranmer GK;Dong Y;Wang YT;Tang JS;Li CF;Walmsley IA;Guo GC
• 通讯作者: Guo GC