Evaluation of Noise and Crosstalk in Neutral Atom Quantum Computers

Jakub Szefer; Konstantinos-Nikolaos Papadopoulos; Pranet Sharma; Yizhuo Tan

REVIEW 4 major objections 6 minor 1 cited by

Evaluation of Noise and Crosstalk in Neutral Atom Quantum Computers

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Co-located neutral-atom simulations are safe again at an 8 µm separation.

desk verdict Real hardware noise maps plus a plausible idea, but the crosstalk and MTD claims are simulator-only and the MTD result is confounded by a missing stationary control. read the letter →

arxiv 2507.22140 v2 pith:GSQJAI53 submitted 2025-07-29 quant-ph cs.CR

Pranet Sharma , Yizhuo Tan , Konstantinos-Nikolaos Papadopoulos , Jakub Szefer This is my paper

classification quant-phcs.CR MSC 81P68 PACS 03.67.-a

keywords neutralatomquantumcomputingcrosstalkanalogHamiltoniansimulationmulti-tenancymovingtargetdefensenoiseevaluationvanderWaalsinteractionsRydbergblockade

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper asks whether separate analog Hamiltonian simulations can run next to each other on a neutral atom quantum computer without corrupting each other. It reports that co-location induces crosstalk that depends on separation: in three-qubit registers built around the platform's default 5.5 µm triangular configuration, the largest fidelity loss occurred at 5 µm separation, while fidelity returned toward the no-co-location baseline beyond 8 µm. It also finds that noise on the physical QPU drifts week to week in a mostly random pattern. To exploit the spacing result without wasting the chip, the paper proposes a Moving Target Defense that physically relocates the victim simulation during execution, and reports near-baseline fidelity (0.995 ± 0.02) when the victim is moved twice by 4 µm. The spatial and defense experiments used a cloud-based analog Hamiltonian simulator because the physical QPU does not yet support local detuning without a specialized access route.

What carries the argument

The load-bearing object is the time-dependent Hamiltonian of Eq. (1), whose all-pairs van der Waals term couples every pair of atoms even when the atoms belong to logically separate simulations. The shifting-field term $H_{\mathrm{shift},k}(t) = -\Delta_{\mathrm{local}}(t) h_k n_k$ creates local detuning, and detuning changes the van der Waals interaction of that atom—this is how co-located simulations crosstalk and how an attacker could amplify disruption. The quantitative workhorse is relative fidelity, the averaged percent difference in final qubit counts from a control, computed on the cloud simulator. The Moving Target Defense is the proposed countermeasure: physically relocate the victim simulation within the optical-tweezer array before or after execution steps so a fixed attacker position cannot hold a damaging separation.

What would settle it

Run the same co-located three-qubit experiment on the physical QPU with local detuning: if relative fidelity at 8 µm separation is not statistically indistinguishable from the no-co-location control, or if moving the victim twice by 4 µm does not restore fidelity near 1, then the claimed spacing bound and the Moving Target Defense fail on hardware.

Watch

Extended reading notes

Core claim

At the center is a three-qubit register forming an equilateral triangle of side 5.5 µm, with a local shifting field detuning one qubit at 20 times the drive frequency; fidelity is scored as relative fidelity of final qubit counts against a control. Repeating the register over a 40 µm by 50 µm grid for four weeks showed temporal noise that is mostly random and consistent with thermal drift and optical imperfections. Co-locating a second identical simulation and moving it in 1 µm steps showed crosstalk with a clear distance dependence: the largest drop, relative fidelity 0.882 ± 0.010, came at 5 µm separation, and values recovered toward 1 at separations beyond 8 µm, which the paper identifies as a promising lower bound for safe co-location. A Moving Target Defense that shifts the victim simulation twice by 4 µm at 5 µm separation restored relative fidelity to 0.995 ± 0.02, close to the no-attack baseline. The paper concludes that separation distance controls inter-simulation crosstalk and that relocation is a viable runtime defense for multi-tenant neutral atom systems.

Load-bearing premise

The spatial-crosstalk and defense results were produced by the cloud-based analog Hamiltonian simulator rather than the physical QPU, because the QPU does not yet support local detuning without specialized access; the argument assumes the simulator's all-pairs van der Waals crosstalk matches the hardware's crosstalk closely enough for the 8 µm bound and the defense to transfer.

Editorial extensions

If this is right

Cloud operators can treat about 8 µm as a minimum physical separation between co-located analog Hamiltonian jobs for the tested register geometry.
Relocating the victim simulation during execution keeps fidelity near the no-attack level, making Moving Target Defense a candidate runtime protection for multi-tenant neutral atom clouds.
Because temporal noise is mostly random week to week, a spatial separation rule is more dependable than trying to schedule around known noisy regions.
The spatial and defense results can be repeated on the QPU itself once local detuning is available through the specialized access route.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Because the spatial and defense numbers come from the simulator's ideal van der Waals model, the 8 µm bound may shift on real hardware where optical imperfections, imperfect Rydberg blockade, and transport errors add crosstalk and decoherence.
Moving a simulation trades crosstalk for atom-transport errors; on a real QPU the 0.995 fidelity could degrade if the relocation itself disturbs the atoms.
The 8 µm value is tied to the tested three-qubit geometry, detuning strength, and density; other register shapes and atom counts likely move the threshold, so the rule should be re-measured per workload.
An attacker who controls atom positions could respond by tracking or herding a moving victim, so a practical defense should randomize the relocation pattern rather than use a fixed schedule.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

What should you know? The real data in this paper are the four weekly noise maps on Aquila — those are genuine and useful. Everything about spatial crosstalk and the moving target defense is simulator-only, and the central MTD claim is confounded by a missing control. The paper is honest enough to footnote the simulator limitation, but the abstract overstates what has been shown.

What's new: the temporal noise characterization over a month on a real neutral-atom QPU (Fig. 1) is a small but real contribution. Applying the classical moving target defense idea to co-located neutral-atom simulations is a new angle in the security literature. The paper is clearly written and the idea of an 8 µm separation rule is concrete. The authors also cite the relevant prior work and do not hide the fact that Aquila lacks local detuning without Braket Direct.

Where it falls down: the spatial crosstalk experiments run on the AWS Braket AHS simulator, whose Hamiltonian (Eq. 1) sums van der Waals interactions over every pair of atoms, including pairs across co-located simulations. So the qualitative finding — closer simulations have more crosstalk — is entailed by the input model. The experiment cannot fail to show proximity-dependent crosstalk. That doesn't make the result false, but it makes it a model prediction rather than a hardware measurement. Second, the MTD experiment compares no-defense at 5 µm separation with defense that ends at 13 µm, and there is no stationary control at 13 µm. Table I already shows fidelity 0.984 at 7 µm, so the high 0.995 fidelity in the defense arm very plausibly just reflects final separation. The data do not demonstrate that moving helps; they show at most that greater separation helps, which is the same as Section IV. Third, the quantitative 8 µm bound rests on four data points, one of which (4 µm vs 5 µm) is non-monotonic. That's thin. And because the defense evaluation is on a simulator, it ignores atom-transport errors, imperfect Rydberg blockade, and optical imperfections that would come with physical movement on real hardware.

Net: this is a reasonable first engineering look at co-located neutral-atom AHS, with one genuine hardware dataset and an interesting idea. It is not a demonstrated mitigation. The MTD claim needs a stationary control at final separation, more separation points, and ideally Braket Direct hardware runs before it can be called viable.

Who it's for: people thinking about multi-tenant neutral-atom quantum cloud security. A serious referee should engage with it; the paper is not a desk reject. Send it to review, but be ready for a heavy revision.

Referee Report

4 major / 6 minor

Summary. The paper studies noise and crosstalk in neutral atom quantum computers, using QuEra's Aquila QPU and the AWS Braket AHS simulator. It first reports a temporal noise evaluation on Aquila, mapping relative fidelity over a 40 x 50 micrometer grid weekly for four weeks and concluding that the noise appears mostly random. It then investigates spatial co-location by running two three-qubit simulations at varying separations on the simulator, reporting that fidelity degrades most at 5 micrometers and recovers beyond about 8 micrometers. Finally, it proposes a moving target defense (MTD) that physically relocates the victim simulation during execution, and claims the MTD achieves near-unity relative fidelity and is a viable technique for safe co-location. The central quantitative claims for crosstalk and MTD are based entirely on simulator experiments, while the only QPU data are the temporal noise measurements.

Significance. The temporal noise study on real hardware is a useful, if preliminary, characterization of drift on a neutral atom QPU over a month, and the paper is honest about the lack of local-detuning support on Aquila. If the spatial crosstalk and MTD results were validated on hardware, they would be relevant to multi-tenant neutral atom quantum computing, a topic with little published work. Credit is due for collecting repeated real-device measurements and for clearly documenting the experimental setup. However, the load-bearing spatial results are currently only simulator demonstrations, and as I detail below, they do not yet establish the paper's main claims about crosstalk or the MTD as a hardware-relevant defense.

major comments (4)

[IV, Eq. (1), Fig. 2, Table I] The spatial crosstalk experiment is conducted on the AWS Braket AHS simulator, whose Hamiltonian is Eq. (1). That equation sums van der Waals interactions over every pair of atoms, including pairs belonging to different, co-located simulations. Therefore, proximity-dependent crosstalk is built into the model by construction; the experiment cannot fail to find it. To support the claim that close proximity of concurrent simulations increases crosstalk, the authors need to either validate the simulator result on Aquila (e.g., via Braket Direct) or present a control that demonstrates the effect is not a trivial consequence of Eq. (1), such as running the same geometry with inter-simulation couplings removed. As written, the central finding in Section IV is a restatement of the input model rather than a measurement.
[V, Fig. 3, Fig. 4] The MTD experiment is missing the necessary control. The 'Defense' arm moves the victim from 5 micrometers to 13 micrometers, while the 'No Defense' arm keeps it at 5 micrometers. Table I already shows relative fidelity of 0.984 +/- 0.009 at only 7 micrometers, so the high fidelity of the defense arm (0.995 +/- 0.02) is plausibly explained entirely by the larger final separation. Without a control where the victim remains stationary at 13 micrometers, or where the same net displacement is applied in a single step, the data do not demonstrate that moving the simulation mitigates crosstalk; they only reproduce the distance-dependence already claimed in Section IV. The conclusion in Section V that the MTD 'can mitigate the impact of noise and crosstalk' is therefore unsupported as stated.
[II.B, V, footnote 1, Abstract] All spatial crosstalk and MTD results come from the AWS Braket AHS simulator, which implements the ideal Hamiltonian of Eq. (1). As the paper notes, Aquila does not support local detuning without specialized Braket Direct access, so the QPU experiments do not exercise the detuning field that drives the crosstalk mechanism. The simulator does not model optical imperfections, imperfect Rydberg blockade, or the atom transport errors that the MTD would introduce on real hardware. Yet the abstract and conclusion state, without qualification, that the findings demonstrate crosstalk and that the MTD is viable for enabling safe and reliable co-location on neutral atom quantum hardware. These hardware-level claims are not justified by the evidence; at most, the results are a model-based prediction that needs QPU validation.
[II.B, III, IV] The paper's key quantitative metric, 'relative fidelity,' is defined as the averaged percent difference in final qubit counts relative to a control. This is not a fidelity in the usual quantum-information sense, and because final atom counts are integer shot outcomes, the metric conflates statistical sampling noise with true state fidelity. The paper does not report the number of shots per data point, the variance across runs, or how the control was calibrated, which makes it difficult to assess whether the differences in Table I and Figure 1 are statistically significant. Given that this metric underlies all quantitative conclusions, the authors should either replace it with a standard state-fidelity estimate or provide a careful justification and full statistical reporting.

minor comments (6)

[IV, Fig. 2] The text says 'Euler distance' but should say 'Euclidean distance.'
[III, Fig. 1] The heatmap colorbars lack labels and units, and the axes are labeled only with distances. Adding 'Relative Fidelity' and specifying the colormap range would improve interpretability.
[IV, Table I] The data are non-monotonic: the 5 micrometer offset has lower fidelity than the 4 micrometer adjacent configuration. The text describes the trend as 'roughly linear' for offsets of 5 micrometers or more, but Table I contains too few points to support a linear fit, and the text should acknowledge the variability explicitly.
[V, Fig. 3, text] The text says the victim was moved 'twice in 4 micrometer intervals,' but Figure 3 appears to show three movement steps. Please reconcile the figure with the text and describe the schedule precisely.
[II.B] When introducing Delta_local, the paper says it is '20 times the frequency of the uniform driving field,' but Delta_local has units of rad/sec, so it is likely the authors mean 20 times the amplitude or angular frequency. Please clarify the wording.
[II.A, Eq. (2)] In Eq. (2), the symbols h_k and n_k are not defined before use. The paper later refers to 'atom-dependent pattern' and the number operator, but these should be defined explicitly at the point of introduction.

Circularity Check

2 steps flagged · score 6.0 of 10

Spatial crosstalk result restates the simulator's all-pairs van der Waals Hamiltonian; the MTD claim is confounded with final separation.

self definitional [Section II.A (Eq. 1, and text after Eq. 2); Section IV (Spatial Noise Evaluation)]
"the van der Waals field is internal to the simulation and governs the interactions between each pair of qubits [1] ... The impact of the shifting field creates an effect on the qubit known as "detuning" – shifting its frequency away from resonance – which changes the van der Waals interactions of that qubit, thereby inducing a form of crosstalk [7]."

The spatial crosstalk experiment was run on the AWS Braket AHS simulator, which implements Eq. (1). That Hamiltonian's last sum includes HvdW,j,k for every pair of atoms in the register, including pairs belonging to different co-located simulations (qsV and qsA). The paper itself defines crosstalk as the consequence of detuning-modified van der Waals interactions. Moving the simulations apart only changes the 1/r^6 couplings already present in the simulator's model, so the observed fidelity-vs-distance curve — and the 8 µm bound — is a direct output of the input Hamiltonian, not an independent hardware-derived discovery.
renaming known result [Section V (Moving Target Defense); compare Table I (Offset 3)]
"Within our moving target defense experiment, we ran qsA and qsV with a separation of 5 µm for "No Defense" data, and moved qsV away from qsA twice in 4 µm intervals for the "Defense" data. ... with the implementation of a moving target defense, qsV had a relative fidelity of 0.995 ± 0.02, indicating close to expected performance"

The defense arm begins at 5 µm and ends at 13 µm, while the no-defense arm is stationary at 5 µm. Table I already reports that a stationary 7 µm separation yields 0.984 ± 0.009, statistically close to the defense arm's 0.995 ± 0.02. There is no control with qsV stationary at the final 13 µm separation, so the higher fidelity in the defense arm is fully explained by the final separation — the same spatial-separation effect characterized in Section IV. The conclusion 'a moving target defense can mitigate the impact of noise and crosstalk' therefore renames the known distance dependence as a defensive technique without isolating movement as the causal factor.

full rationale

The paper's central qualitative crosstalk claim is self-definitional in the simulator context: Eq. (1) sums van der Waals interactions over all atom pairs, and the paper identifies detuning-modified van der Waals interactions as crosstalk. The co-located spatial scan measures exactly this built-in coupling, so the observed distance dependence is an output of the model rather than an independent empirical result. The MTD claim is additionally confounded: the defense manipulation is a movement that also changes final separation, and the observed fidelity is indistinguishable from the already-measured stationary separation effect; no control at stationary 13 µm is provided. The temporal noise evaluation on Aquila is independent and not circular, and the paper contains no load-bearing self-citation of a uniqueness theorem. However, the spatial and MTD conclusions—the paper's chief novel findings—reduce to the input Hamiltonian and to the previously established separation effect, respectively, warranting a score of 6 (partial circularity).

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The engine behind the crosstalk results is a simulator whose Hamiltonian (Eq. 1) already includes the all-pairs van der Waals coupling that produces the interference, so the co-location experiments carry no independent hardware evidence. The hand-chosen values (Δlocal, register geometry, sampling grid, MTD schedule) shape all quantitative outputs. No new physical entities are postulated; the moving target defense is a scheduling strategy, not an invention of new physics.

free parameters (4)

Local detuning Δlocal = 5×10^7 rad/s (20x the driving frequency)
Hand-chosen in Section II.B to detune the top qubit and define the 'attack' simulation; all crosstalk and MTD results depend on this value.
Register geometry = Equilateral triangle, side 5.5 µm
Taken as QuEra's default starting configuration (Section II.B); sets the length scale for all co-location distances.
Spatial sampling grid = 10 µm steps over a 40 by 50 µm area
Chosen as roughly double QuEra's default spacing (Section III); noise-map conclusions are conditional on this sampling.
MTD movement schedule = two 4 µm moves: 5 µm to 13 µm separation
Hand-chosen in Section V; only this one schedule is tested, so defense efficacy is demonstrated for this schedule alone.

assumptions (3)

domain assumption The AWS Braket AHS simulator faithfully implements Eq. (1), including all-pairs van der Waals couplings, and its crosstalk behavior represents real hardware crosstalk.
Invoked in Sections IV and V; all spatial crosstalk and moving target defense results are produced by the simulator because Aquila does not support local detuning (Section II.B, Section V footnote 1).
domain assumption A detuned qubit induces crosstalk in nearby qubits through modified van der Waals interactions.
Section II.A states this, citing [7]; the attack model and the interpretation of Table I depend on it.
ad hoc to paper Relative fidelity, defined as the averaged percent difference between measured and control qubit counts, is a valid fidelity proxy for analog Hamiltonian simulations.
Defined in Section II.B and used for all conclusions; no validation of this metric against a standard fidelity measure is provided.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Evaluation of Noise and Crosstalk in Neutral Atom Quantum Computers." pith.science (2026). https://pith.science/paper/GSQJAI53

@misc{pith2026250722140,
  author       = {Pith},
  title        = {Pith review of: Evaluation of Noise and Crosstalk in Neutral Atom Quantum Computers},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GSQJAI53}},
  note         = {Machine review of arXiv:2507.22140}
}

read the original abstract

This work explores and evaluates noise and crosstalk in neutral atom quantum computers. Neutral atom quantum computers are a promising platform for analog Hamiltonian simulations, which rely on a sequence of time-dependent Hamiltonians to model the dynamics of a larger system and are particularly useful for problems in optimization, physics, and molecular dynamics. However, the viability of running multiple simulations in a co-located or multi-tenant environment is limited by noise and crosstalk. This work conducts an analysis of how noise faced by simulations changes over time, and investigates the effects of spatial co-location on simulation fidelity. Findings of this work demonstrate that the close proximity of concurrent simulations can increase crosstalk between them. To mitigate this issue, a Moving Target Defense (MTD) strategy is proposed and evaluated. The results confirm that the MTD is a viable technique for enabling safe and reliable co-location of simulations on neutral atom quantum hardware.

Figures

Figures reproduced from arXiv: 2507.22140 by the authors.

**Figure 1.** Temporal noise evaluation on the Aquila quantum computer over period of four weeks in May, 2025. The darker green colors represent parts of the array that performed more closely to the determined control. Uniform coloring – such as visible in Fig 1d – implies less variance in final qubit counts across multiple runs, while non-uniform coloring – such as visible in Fig 1a – implies more variance. The shifting field ca… view at source ↗

**Figure 2.** Spatial noise evaluation settings tested used this work. “Adjacent” configuration set the simulations 4 µm apart. “Offset 1” set the qubits 5 µm apart, “Offset 2” set the qubits 6 µm apart, and “Offset 3” set the qubits 7 µm apart. The figure is not to scale. The diagonal Euler distance between the qubits was used as the independent variable to control for horizontal or vertical biases in crosstalk. TABLE I: Simulat… view at source ↗

**Figure 3.** Moving target defense overview. The figure is not to scale. At each step, the victim circuit was moved 4 µm. The simulation can optionally be moved further after measurement. No Defense Defense 0.0 0.2 0.4 0.6 0.8 1.0 Sim. Fidelity Relative to No Attack [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: Fidelity without and with moving target defense. Initially, the separation between victim and attacker was 5 µm. electronics in neutral atom architectures, and this could be a valuable future direction. VII. CONCLUSION This work investigated noise and crosstalk in neut…

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

Crosstalk In Contemporary Quantum Devices
quant-ph 2026-05 unverdicted novelty 1.0 of 10

Review synthesizing crosstalk mechanisms, mitigation strategies, and security vulnerabilities across major quantum computing platforms from existing literature.

Reference graph

Works this paper leans on

17 extracted references · 12 canonical work pages · cited by 1 Pith paper

[1]

Aquila: Quera’s 256-qubit neutral-atom quantum computer,

J. Wurtz, A. Bylinskii, B. Braverman, J. Amato-Grill, S. H. Cantu, F. Huber, A. Lukin, F. Liu, P. Weinberg, J. Long et al. , “Aquila: Quera’s 256-qubit neutral-atom quantum computer,” arXiv preprint arXiv:2306.11727, 2023

arXiv 2023
[2]

Surface code quantum computing by lattice surgery,

D. Horsman, A. G. Fowler, S. Devitt, and R. Van Meter, “Surface code quantum computing by lattice surgery,” New Journal of Physics, vol. 14, no. 12, p. 123011, 2012

2012
[3]

Practical quantum advantage in quantum simulation,

A. J. Daley, I. Bloch, C. Kokail, S. Flannigan, N. Pearson, M. Troyer, and P. Zoller, “Practical quantum advantage in quantum simulation,” Nature, vol. 607, no. 7920, pp. 667–676, 2022

2022
[4]

Universal quantum hamiltonians,

T. S. Cubitt, A. Montanaro, and S. Piddock, “Universal quantum hamiltonians,” Proceedings of the National Academy of Sciences , vol. 115, no. 38, pp. 9497–9502, 2018

work page 2018
[5]

Quantum advantage and stability to errors in analogue quantum simulators,

R. Trivedi, A. Franco Rubio, and J. I. Cirac, “Quantum advantage and stability to errors in analogue quantum simulators,” Nature Communi- cations, vol. 15, no. 1, p. 6507, 2024

work page 2024
[6]

Quantum cloud computing: A review, open problems, and future directions,

H. T. Nguyen, P. Krishnan, D. Krishnaswamy, M. Usman, and R. Buyya, “Quantum cloud computing: A review, open problems, and future directions,” arXiv preprint arXiv:2404.11420 , 2024

arXiv 2024
[7]

Quantum computing with neutral atoms,

D. S. Weiss and M. Saffman, “Quantum computing with neutral atoms,” Physics Today, vol. 70, no. 7, pp. 44–50, 2017

work page 2017
[8]

Amazon Braket: Quantum computing on AWS,

Amazon Web Services, “Amazon Braket: Quantum computing on AWS,” https://aws.amazon.com/braket/, 2025

work page 2025

Show all 17 references

[9]

qbraid lab user guide,

qBraid, “qbraid lab user guide,” https://lab.qbraid.com/, 2025

2025
[10]

Amazon braket features (with braket direct),

Amazon Web Services, “Amazon braket features (with braket direct),” https://aws.amazon.com/braket/features/#Braket Direct, Jul 2025

2025
[11]

Neutral atom quantum computing hardware: performance and end-user perspective,

K. Wintersperger, F. Dommert, T. Ehmer, A. Hoursanov, J. Klepsch, W. Mauerer, G. Reuber, T. Strohm, M. Yin, and S. Luber, “Neutral atom quantum computing hardware: performance and end-user perspective,” EPJ Quantum Technology, vol. 10, no. 1, p. 32, 2023

2023
[12]

Toward proactive, adaptive defense: A survey on moving target defense,

J.-H. Cho, D. P. Sharma, H. Alavizadeh, S. Yoon, N. Ben-Asher, T. J. Moore, D. S. Kim, H. Lim, and F. F. Nelson, “Toward proactive, adaptive defense: A survey on moving target defense,” IEEE Communications Surveys & Tutorials, vol. 22, no. 1, pp. 709–745, 2020

2020
[13]

Benchmarking a neutral-atom quantum computer,

N. Wagner, C. Poole, T. Graham, and M. Saffman, “Benchmarking a neutral-atom quantum computer,” International Journal of Quantum Information, vol. 22, no. 04, p. 2450001, 2024

2024
[14]

Qubithammer attacks: Qubit flipping attacks in multi-tenant superconducting quantum computers,

Y . Tan, N. Choudhury, K. Basu, and J. Szefer, “Qubithammer attacks: Qubit flipping attacks in multi-tenant superconducting quantum computers,” 2025. [Online]. Available: https://arxiv.org/abs/2504.07875

2025 arXiv
[15]

Hacking quantum computers with row hammer attack,

F. Almaguer-Angeles, P. R. Dieguez, A. S. H., and M. Pawłowski, “Hacking quantum computers with row hammer attack,” 2025. [Online]. Available: https://arxiv.org/abs/2503.21650

2025 arXiv
[16]

Short paper: Device- and locality-specific fingerprinting of shared nisq quantum computers,

A. Mi, S. Deng, and J. Szefer, “Short paper: Device- and locality-specific fingerprinting of shared nisq quantum computers,” in Hardware and Architectural Support for Security and Privacy , 2021

2021
[17]

Exploration of power side-channel vulnerabilities in quantum computer controllers,

C. Xu, F. Erata, and J. Szefer, “Exploration of power side-channel vulnerabilities in quantum computer controllers,” in Conference on Computer and Communications Security , 2023

2023

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [1]

Aquila: Quera’s 256-qubit neutral-atom quantum computer,

J. Wurtz, A. Bylinskii, B. Braverman, J. Amato-Grill, S. H. Cantu, F. Huber, A. Lukin, F. Liu, P. Weinberg, J. Long et al. , “Aquila: Quera’s 256-qubit neutral-atom quantum computer,” arXiv preprint arXiv:2306.11727, 2023

arXiv 2023

[2] [2]

Surface code quantum computing by lattice surgery,

D. Horsman, A. G. Fowler, S. Devitt, and R. Van Meter, “Surface code quantum computing by lattice surgery,” New Journal of Physics, vol. 14, no. 12, p. 123011, 2012

2012

[3] [3]

Practical quantum advantage in quantum simulation,

A. J. Daley, I. Bloch, C. Kokail, S. Flannigan, N. Pearson, M. Troyer, and P. Zoller, “Practical quantum advantage in quantum simulation,” Nature, vol. 607, no. 7920, pp. 667–676, 2022

2022

[4] [4]

Universal quantum hamiltonians,

T. S. Cubitt, A. Montanaro, and S. Piddock, “Universal quantum hamiltonians,” Proceedings of the National Academy of Sciences , vol. 115, no. 38, pp. 9497–9502, 2018

work page 2018

[5] [5]

Quantum advantage and stability to errors in analogue quantum simulators,

R. Trivedi, A. Franco Rubio, and J. I. Cirac, “Quantum advantage and stability to errors in analogue quantum simulators,” Nature Communi- cations, vol. 15, no. 1, p. 6507, 2024

work page 2024

[6] [6]

Quantum cloud computing: A review, open problems, and future directions,

H. T. Nguyen, P. Krishnan, D. Krishnaswamy, M. Usman, and R. Buyya, “Quantum cloud computing: A review, open problems, and future directions,” arXiv preprint arXiv:2404.11420 , 2024

arXiv 2024

[7] [7]

Quantum computing with neutral atoms,

D. S. Weiss and M. Saffman, “Quantum computing with neutral atoms,” Physics Today, vol. 70, no. 7, pp. 44–50, 2017

work page 2017

[8] [8]

Amazon Braket: Quantum computing on AWS,

Amazon Web Services, “Amazon Braket: Quantum computing on AWS,” https://aws.amazon.com/braket/, 2025

work page 2025