Quantum computing with atomic qubit arrays: confronting the cost of connectivity
M. Saffman \orcidID 0000-0001-6398-2097 · 2025
Read Quantum computing with atomic qubit arrays: confronting the cost of connectivity by M. Saffman \orcidID 0000-0001-6398-2097 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-11218.txt.
Introduction
Saffman reviews quantum computing with neutral atom arrays in this 2025 paper. The approach has leapt ahead in recent years, yet scaling to useful sizes hits a hard wall: linking far-apart qubits fast enough. Atom arrays carry two linking tools, long-range Rydberg gates and physical atom motion.
The review weighs prototype layouts by how fast they connect large entangled circuits. It also scores them on cycle speed for error-correcting codes based on measurement. Mixing several atom species emerges as a promising route for that speed. The author frames connectivity cost as the field's central bill to pay.
Catalog record
| CREATOR | M. Saffman \orcidID 0000-0001-6398-2097 |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 62 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-11218
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Quantum computing with atomic qubit arrays: confronting the cost of connectivity M. Saffman \orcidID 0000-0001-6398-2097 Abstract These notes present a review of the status of quantum computing with arrays of neutral atom qubits, an approach which has demonstrated remarkable progress in the last few years. Scaling digital quantum computing to qubit counts and control fidelities that will enable solving outstanding scientific questions, and provide commercial value, is an outstanding challenge, not least because of the requirement of connecting and entangling distant qubits. Long-range Rydberg gates and physical motion outfit atomic qubit arrays with tools for establishing connectivity. These tools operate on different timescales and with distinct levels of parallelization. We analyze several prototypical architectures from the perspective of achieving fast connectivity for circuits with large scale entanglement, as well as fast cycle times for measurement based quantum error correcting codes. Extending Rydberg interactions to multiple atomic species has emerged as a promising route to achieving this latter requirement. 1 Introduction The concept of a quantum computer is not new [ 1 , 2 , 3 , 4 ] , yet the timeline for wide availability of quantum computers that are useful for solving problems in science or commerce is uncertain. 1 1 1 The material presented here is based on lectures given at the International School of Physics “Enrico Fermi” in Varenna, Italy in July, 2024. Given the availability of other review papers and the very rapid pace of development of quantum computing with neutral atoms I have chosen not to include all the background material presented in the lectures. Instead some new developments up until summer 2025 when these notes were finalized are presented. The uncertainty does not arise from questions about the predictive power of quantum mechanics, which was established 100 years ago and has been verified in countless experiments. Rather, scientists and engineers who are striving to build quantum computers do not yet have a verified blueprint for how to proceed. It should also be acknowledged that objections to the completeness of quantum mechanics [ 5 ] , and its seemingly absurd predictions about the physical world [ 6 , 7 , 8 ] have persisted [ 9 ] , despite the success of experiments which have confirmed the violation of Bell’s inequality [ 10 , 11 , 12 ] as predicted by quantum theory [ 13 ] . One of the fundamental applications of large scale quantum computing may well be the creation of new tools for studying the foundations of quantum theory. There is no single reason why developing a quantum computer is so challenging. Even so, a few obstacles stand out. Quantum computational advantage derives from the possibility of controlling quantum phenomena including superposition and entanglement. Preserving these quantum phenomena demands that the qubits are well isolated from the environment. On the other hand a computer is not just a science experiment but is a machine that can be programmed to perform desired tasks, which requires control. Achieving both isolation and control in a single system is difficult and has been tackled most successfully in just a handful of physical platforms, including trapped ions, superconducting circuits, photonics, electrons in low-dimensional semiconductors, and neutral atoms [ 14 ] . Apart from the dichotomy of isolation and control there is the challenge of connectivity. In conventional computers information is freely copied and moved across circuits on wires. In the quantum world unknown states cannot be copied [ 15 , 16 ] , and can only be moved to a different physical location if no trace of the state is left behind. As a consequence performing logic with qubits that may be physically far apart is not trivial and significantly adds to the challenge of scaling up quantum computers. This challenge is present in all of the aforementioned platforms. Neutral atoms provide two mechanisms for achieving connectivity: long-range Rydberg interactions [ 17 , 18 ] , and physical transport of qubits [ 19 ] . These two mechanisms operate on different spatial and temporal scales which has consequences for how to optimize the connectivity architecture. Figure 1: A bifurcated scenario for the development of quantum computers beyond the current NISQ era. High performance NISQ (HP-NISQ) for beyond classical capabilities requires moderate increases in qubit count and a substantial increase in gate fidelity. Quantum error correction and Fault Tolerant Quantum Computing(FTQC) will require orders of magnitude more qubits and a moderate increase in gate fidelity. Before proceeding to specifics of neutral atom qubits, it seems useful to delineate two distinct paths forward, and the implications for future progress. These paths are indicated in Fig. 1 . The current state of play has been termed the NISQ era for Noisy Intermediate Scale Quantum [ 20 ] . Working quantum computers exist and have demonstrated impressive performance, even though they have a limited number of qubits and suffer from errors in control and measurements. Despite these limitations NISQ machines have performed calculations based on random circuit sampling (RCS) that produce bitstrings with a statistical distribution that cannot be mimicked by conventional computers. These so-called “Quantum Supremacy" experiments [ 21 , 22 , 23 , 24 , 25 ] demonstrate the capability of performing a particular calculation that is beyond the reach of conventional computers. but not yet a calculation that has wider utility. The possibility of using RCS not just to demonstrate the power of a quantum computer, but to solve useful problems, has not been decisively demonstrated, although there are promising opportunities involving applications of certified randomness [ 26 ] and quantum fingerprinting for comparing remote databases [ 27 ] . Other demonstrations of NISQ hardware that seek to demonstrate computational advantage for digital quantum circuits have targeted the transverse field Ising model [ 28 , 29 ] which has broad applicability as a description of interactions and dynamics of many-body spin systems. Very recent results enabled by high fidelity entangling gates simulated prethermalization in a regime beyond currently available classical techniques [ 29 ] . While the progress in RCS and transverse field Ising simulations with quantum processors is remarkable, the progress has also inspired development of classical algorithms that successfully competed with early quantum demonstrations [ 30 , 31 ] . This situation with a back and forth competition between conventional computers and digital quantum hardware begs the question of the viability of quantum computational advantage in the current NISQ era without quantum error correction and fault tolerance. To clarify this question we need to establish a connection between the fidelity of expectation values of observables provided by a quantum circuit and the classical cost of simulating the observable. The performance of state of the art NISQ processors is typically limited by the fidelity of the two-qubit gate used for entanglement. As discussed in [ 32 ] the expectation value of an operator \hat{O} that is obtained from a noisy quantum state with density matrix \rho can be expressed as {\rm tr}(\rho\hat{O})={\mathcal{F}}_{\rm eff}\langle\hat{O}\rangle_{\rm ideal} where {\mathcal{F}}_{\rm eff} is the effective fidelity of the observable and \langle\hat{O}\rangle_{\rm ideal} is the noise free expectation value. An estimate for the fidelity is {\mathcal{F}}_{\rm eff}\sim e^{-\epsilon V_{\rm eff}} (1) where \epsilon is the two-qubit gate infidelity and V_{\rm eff} is the