Atoms Reordered into Perfect Arrays in a Fraction of the Usual Time

A new planning algorithm reduces the time needed to create defect-free atom arrays in neutral-atom systems. Koki Aoyama and colleagues at the The University of Osaka in collaboration with National Institutes of Natural Sciences and The Graduate University for Advanced Studies have developed an approach that achieves a reconfiguration plan with a transportation cost scaling with the square root of the number of atoms, representing a key improvement over existing techniques. Numerical simulations using a 632×632 atom array demonstrate a reduction in total transportation cost to one-seventh of current state-of-the-art algorithms, alongside a 32% to 35% increase in atom captures, enabling scalable neutral-atom quantum computation

Efficient atom array reconfiguration via decomposed one-dimensional shuttling tasks

A 632×632 atom array saw its total transportation cost for atom reconfiguration reduced to one-seventh of that achieved by state-of-the-art algorithms. Previous methods struggled to efficiently manage the complexity of atom movement beyond relatively small arrays, and scaling beyond this point was considered impractical due to the exponential increase in computational demands. The core challenge in neutral atom array reconfiguration lies in the combinatorial complexity of moving each atom from its initial position to a desired target position without collisions or significant loss of coherence. Traditional algorithms often treat each atom’s movement independently, leading to a computational bottleneck as the number of atoms, N, increases. This new algorithm achieves its efficiency by decomposing complex rearrangements into a maximum of three one-dimensional ‘shuttling’ tasks, enabling parallel atom transport via a two-dimensional lattice pattern generated by acousto-optic deflectors, devices that steer atoms using light. Acousto-optic deflectors (AODs) create spatially varying optical potentials by diffracting laser beams with radio frequency signals, allowing for precise and dynamic control over atom positions.

Simulations revealed a 32 to 35 percent increase in the number of atoms successfully captured during reconfiguration. This improvement in capture rate is crucial for maintaining the fidelity of quantum computations, as atom loss directly translates to errors. The algorithm employs the Gale-Ryser theorem to reliably achieve desired atomic arrangements, and a ‘peephole optimisation’ technique further enhances efficiency for grid-like target geometries. The Gale-Ryser theorem, a result from combinatorial optimisation, guarantees the existence of a solution for dividing resources, in this case, the movement capacity of the AODs, to satisfy the demands of each atom’s relocation. Peephole optimisation involves locally refining the transport plan by examining small ‘peepholes’ of the array to identify and eliminate redundant or inefficient movements. When using a 632×632 atom array, this new algorithm reduces total transportation cost to one-seventh of existing methods. Breaking down complex atom movements into a maximum of three one-dimensional ‘shuttling’ tasks underpins this improvement, allowing for parallel transport across a two-dimensional lattice. The algorithm’s time complexity scales as $\mathcal{O}(\sqrt N)$, a significant advantage over the $\mathcal{O}(N)$ or higher scaling of many conventional approaches. These results suggest a strong step towards scalable neutral-atom quantum computers, although the current model assumes ideal conditions and does not yet account for practical limitations such as diffraction constraints or potential atom loss during transport. Acknowledging that perfect resource divisibility is an idealisation, this remains a significant step forward for neutral atom quantum computing. The reduction in transportation cost directly translates to reduced experimental time and energy expenditure, making larger-scale quantum simulations more feasible.

Algorithm efficiency versus physical limitations in neutral atom arrays

Increasingly precise control over individual atoms is demanded by the creation of stable and scalable quantum computers, but this new algorithm, demonstrably efficient in simulation, relies heavily on the Gale-Ryser theorem to guarantee solutions. The theorem assumes perfect divisibility of resources, a condition rarely met in physical systems, as real-world atom manipulation suffers from imperfections like diffraction and atom loss. Diffraction, caused by the wave-like nature of light, limits the precision with which AODs can focus laser beams, potentially leading to unintended atom displacements. Atom loss can occur due to collisions with background gas molecules or imperfections in the trapping potential. This raises a key tension: will the mathematical elegance of the algorithm overcome the messy realities of experimental implementation, or will unavoidable errors accumulate and undermine its promised scalability. The sensitivity of neutral atom qubits to environmental noise necessitates extremely high vacuum conditions and precise control over laser frequencies and intensities, adding further complexity to experimental realisation.

A substantial margin of reduction in fundamental transportation cost, to one-seventh of current methods, provides an important buffer against these imperfections. This new planning algorithm streamlines the arrangement of atoms in neutral-atom systems, offering a significant advance for building scalable quantum computers. By dividing complex rearrangements into a maximum of three simple, one-dimensional movements, termed ‘shuttling’, the technique enables parallel atom transport using acousto-optic deflectors, devices that steer atoms with light. Simulations utilising a 632×632 atom array demonstrated a reconfiguration time proportional to the square root of the number of atoms, alongside a substantial reduction in the energy required to move atoms. The energy reduction is achieved by minimising the total distance travelled by all atoms during reconfiguration. Furthermore, the algorithm’s ability to decompose complex movements into simpler shuttling tasks simplifies the control signals required for the AODs, potentially reducing the complexity of the experimental setup. Future work will focus on incorporating realistic noise models and error correction strategies into the simulations to assess the algorithm’s robustness in the presence of imperfections. Investigating the impact of diffraction limits and atom loss rates on the overall performance is also crucial for guiding experimental implementation. The development of more sophisticated error mitigation techniques, tailored to the specific characteristics of neutral atom systems, will be essential for realising the full potential of this new planning algorithm and achieving truly scalable quantum computation.

The research demonstrated a new planning algorithm capable of arranging up to 632×632 atoms with a reconfiguration time proportional to the square root of the number of atoms. This represents an improvement in scalability for neutral-atom systems, as the algorithm reduced total transportation cost to one-seventh of existing methods and increased atom captures by 32 to 35 percent. The algorithm achieves this by breaking down complex atom rearrangements into simpler, parallel ‘shuttling’ tasks using acousto-optic deflectors. Researchers intend to further assess the algorithm’s robustness by incorporating realistic noise models and error correction strategies into future simulations.