Quantum Error Correction in a dual-species Rydberg array (QuERy)

量子计算提供了一种新的范式，可以使用在量子力学定律下运行的系统来解决复杂的问题，例如用于增强药物设计的量子化学，或用于设计航空航天和工程新材料的相关介质的建模，这些问题目前超出了现代数字计算机的能力。量子硬件还可以大幅加速计算代价高昂的问题，从与物流相关的经典优化(例如旅行推销员类型的问题)到因式分解。然而，利用量子计算的优势来解决这些现实世界的问题的主要障碍是建立一个能够提供足够数量的高质量、低噪声量子比特(QU比特)的系统所面临的技术挑战。最近的进展见证了许多不同技术的发展来应对这一挑战，包括超导量子比特和囚禁离子，然而，由单独捕获的原子组成的可重构阵列已经成为一种竞争激烈的方法，能够在不损失性能的情况下扩展到大量相同的量子比特。这些目前只有几百个量子比特的系统提供了探索量子计算早期好处的机会，但受到噪声和错误的限制，这限制了系统的性能和解决有用问题的能力。为了克服这个限制，有必要执行纠错来修复错误，否则会破坏计算机的输出。在数字硬件中，这是通过对逻辑寄存器中存储的信息执行测量来实现的。对于量子计算机来说，这要困难得多，因为信息是以脆弱的叠加态编码的，当直接测量时，这会导致信息的丢失。解决这个问题的一种方法是通过利用具有特殊选择的对称属性的拓扑保护态来使用多个量子比特来对单个逻辑量子比特进行编码来执行量子纠错。为了在不删除量子态的情况下检查错误，额外的量子比特(称为附加量子比特)被用于执行对称性的局部测量，之后可以使用标准的门协议来校正错误。虽然已经提出了一系列用于执行量子纠错的编码方案，但到目前为止，这只在超导体和离子的少数量子比特系统上实现，并且由于仅使用单一原子物种时串扰的挑战，尚未在更可扩展的中性原子平台上进行演示。在Query中，我们将为中性原子量子计算开发一个新的双物种平台，以直接解决将量子硬件从100秒扩大到数百万量子比特的主要障碍，即通过在低温环境中集成来延长原子阵列的捕获寿命，以抑制原子损失造成的错误，并通过使用一个物种编码信息和使用第二个物种执行局部Ancilla测量来执行无串扰、状态选择性的局部测量来演示量子纠错。这项研究将为中性原子量子计算提供新的模式，包括通过随机测量实现验证和基准测试，并开发能够利用拓扑编码实现横向门操作的算法。这些结果对于实现大规模、容错的量子计算机至关重要，这是利用量子计算在学术界和工业界的变革性好处所需的，包括用于药物发现的材料设计、物流和量子化学等应用。

Quantum computing provides a new paradigm in which to use systems operating under the laws of quantum mechanics to solve complex problems such as in quantum chemistry for enhanced drug design or modelling of correlated media for designing new materials for aerospace and engineering which are currently beyond the capability of modern digital computers. Quantum hardware can also provide a dramatic speed-up of computationally expensive problems ranging from classical optimisation relevant to logistics (e.g. travelling salesman type problems) to factorisation. The major barriers to exploiting the advantages of quantum computing for these real world problems however are the technological challenges associated with building a system able to provide a sufficient number of high quality, low noise quantum-bits (qu-bits).Recent progress has seen the development of a number of different technologies to address this challenge, including superconducting qubits and trapped ions, however reconfigurable arrays of individually trapped atoms have emerged as a highly competitive approach able to scale to large numbers of identical qubits without loss of performance. These current systems with a few hundred qubits offer opportunities to explore early benefits of quantum computing, but suffer from noise and errors which place a limit on the performance and their ability to address useful problems.To overcome this limitation, it is necessary to perform error correction to fix the errors that otherwise corrupt the computer output. In digital hardware this is achieved by performing measurements on the information stored in logical registers. For a quantum computer this is much harder to implement, as the information is encoded in fragile superposition states which when measured directly results in a loss of information. One route to address this problem is to perform quantum error correction using a number of qubits to encode a single logical qubit by exploiting topologically protected states with specially chosen symmetry properties. To check for errors without erasing the quantum states, additional qubits (known as ancillas) are used to perform local measurements of the symmetries, after which errors can be corrected using standard gate protocols. Whilst a range of encoding schemes have been proposed for performing quantum error correction, this has so far only been implemented on few qubit systems of superconductors and ions and has yet to be demonstrated on the more scalable neutral atom platform due to challenges with cross talk when a only a single atomic species is used. In QuERy we will develop a new dual-species platform for neutral atom quantum computing to directly address major barriers to scaling up quantum hardware from 100s to millions of qubits, namely extending atom array trapping lifetimes through integration in a cryogenic environment to suppress errors due to atom loss, and demonstrating quantum error correction through the ability to perform cross-talk free, state-selective local measurements by using one species for encoding information and a second species to perform local ancilla measurements.This research will provide new modalities for neutral atom quantum computing, including enabling verification and benchmarking through randomised measurements, and develop algorithms able to exploit topological encodings to implement transverse gate operations. These results are crucial for the realisation of large-scale, fault-tolerant quantum computers as required to exploit the transformative benefits of quantum computing across both academia and industry, including applications such as material design, logistics and quantum chemistry for drug discovery.