通过里德堡相互作用纠缠的中性原子阵列系统在近十年来时间内成为了最具潜力的量子模拟和量子计算实验平台之一。它具有可扩展性高、全联通、相干时间长、并行度高等特点，受到广泛关注。 原子重排是将因碰撞阻塞效应而带缺陷的初始装载阵列转换成无缺陷阵列的重要步骤。随着阵列规模增大，原子重排的难度随之增大。为了加速无缺陷阵列重排过程、降低重排过程中的原子损失、提升真空寿命限制下的无缺陷阵列规模，本文介绍了作者设计的基于现场可编程逻辑门阵列（FPGA）的重排控制系统以及名为“俄罗斯方块算法”的并行重排算法。我们通过多重并行性设计来减少重排耗时。首先，原子占据检测、重排算法决策、控制信号生成等步骤在时间上并行执行；其次，一行（列）的原子在空间上并行移动；最后，俄罗斯方块算法无需等待整张图像都被接收，可以在图像传输过程中仅有局部信息的情况下开始重排。高度集成化和多重并行性的设计有效降低了重排的响应延时和总耗时，进而降低原子与真空背景碰撞导致的损失。系统性能测试的结果显示，该系统显著加速了重排过程，展示了在室温下实现超过1000个原子的无缺陷阵列规模的可能性；而在低温环境下，则有望将系统规模扩展到超过10000个原子。 本文进一步介绍了作者参与设计并搭建的基于$^{87}\rm{Rb}$原子的中性原子阵列平台，整套系统包含真空系统、冷却光系统、磁场系统、控制系统、基于声光偏转器（AOD）和空间光调制器（SLM）的光镊阵列系统等组成部分。通过优化真空设计、使用小型化的铷源和玻璃腔等方式，我们实现了89秒的光镊中单原子寿命。结合FPGA重排系统，目前已成功实现将$24\times24$的随机初始装载阵列重排为$15\times 15$的无缺陷阵列。 此外，论文还探讨了在中性原子阵列系统中加速量子比特测量的方法，这对实现线路中操作和量子纠错等技术至关重要。我们提出了一种基于脉冲光的量子态测量方案，与传统的连续光测量方案相比，该方案通过预测原子在脉冲光下的散射行为，能够在保持高散射率的同时降低杂散光噪声，提高信噪比和探测速度。
Over the past decade, neutral atom arrays with Rydberg interactions have emerged as one of the most promising platforms for quantum simulation and quantum computing. Their scalability, all-to-all connectivity, long coherence time and high parallelism have attracted widespread attention. Atom rearrangement is an important step in converting an initially random loaded array with defects due to collisional blockade into a defect-free array. The difficulty of atom rearrangement increases with the array size.To accelerate the assembling of defect-free arrays, reducing atom loss during rearrangement, and enhancing the scale of defect-free arrays limited by vacuum lifetime,we construct a Field-Programmable Gate Array (FPGA) based rearrangement system along with a parallel rearrange algorithm named Tetris algorithm . The total time cost of the rearrangement is first reduced by processing atom detection, atomic occupation analysis, rearrangement strategy formulation, and acousto-optic deflectors driving signal generation in parallel in time. Then, by simultaneously moving multiple atoms in the same row (column), we save rearrangement time by parallelism in space. Meanwhile, the Tetris algorithm does not have to wait for the entire image to be received; it can start rearrangement with only partial information during the image transmission. % The system features multiple parallelisms, including parallel processing among various functional modules, parallel atom movement, and the ability to process rearrangement during image readout.The highly integrated and parallelized design effectively reduces the latency and time cost, thereby minimizing losses due to collisions with the vacuum background. We then benchmark the overall performance for different target geometries, demonstrating a dramatic boost in rearrangement time cost and the potential to scale up defect-free atomic array to 1000 atoms in room-temperature and 10000 atoms in cryogenic environment. We describe the design and construction of a neutral atom array platform based on $^{87}\rm{Rb}$ atoms, including vacuum system, magneto-optical trap system, control system, and tweezer array system based on acousto-optic deflector and spatial light modulator, among others. Through compact vacuum design and utilizing dispenser as rubidium sources and glass cell, a single-atom vacuum lifetime of 89 seconds in optical tweezers is achieved. With the FPGA rearrangement system, we have successfully rearranged a randomly loaded initial array of $24\times24$ to a defect-free array of $15\times 15$. Furthermore, we discuss quantum state measurement in neutral atom arrays, which is crucial for implementing techniques such as mid-circuit operation and quantum error correction.We propose a pulsed-laser-based quantum state measurement scheme. By predicting the scattering behavior of atoms under pulsed laser, this scheme reduces stray light noise and improves signal-to-noise performance and detection speed, while maintaining a high scattering rate.
