Utility of NISQ devices: optimizing experimental parameters for the fabrication of Au atomic junction using gate-based quantum computers

Takumi Kanezashi, Daisuke Tsukayama, Jun-ichi Shirakashi, Tetsuo Shibuya, Hiroshi Imai

#2050 of 2593 · Quantum Physics
Share
Tournament Score
1328±33
10501750
29%
Win Rate
10
Wins
24
Losses
34
Matches
Rating
3.5/ 10
Significance
Rigor
Novelty
Clarity

Abstract

Feedback-controlled electromigration (FCE) enables precise regulation of atomic migration by carefully optimizing multiple experimental parameters. However, manually fine-tuning these parameters poses significant challenges. This study investigated the feasibility of autonomously fabricating Au atomic junctions through gate-based quantum computing using a noisy intermediate-scale quantum (NISQ) device, which effectively approximates solutions to combinatorial optimization problems. We compared the computational accuracy of the NISQ device against a previously reported D-Wave quantum annealer. The results indicate that the NISQ device achieved lower residual energies and produced higher-quality approximate solutions for large-scale problems than the quantum annealing system.

AI Impact Assessments

(3 models)

Scientific Impact Assessment

Core Contribution

This paper investigates the use of gate-based NISQ quantum computers (IBM Eagle processors) for optimizing feedback-controlled electromigration (FCE) experimental parameters, specifically the feedback voltage (V_FB) scheduling for fabricating Au atomic junctions. The core novelty lies in comparing gate-based quantum computing (via VQE with a hardware-efficient RY-gate Ansatz) against previously reported D-Wave quantum annealers for this specific combinatorial optimization task. The main finding is that NISQ gate-based devices achieve lower residual energies than quantum annealing systems, particularly for larger problem instances, attributed to the fact that gate-based devices don't require the qubit-expensive minor embedding procedure.

Methodological Rigor

The methodology has several notable aspects but also significant concerns:

Strengths in approach:

  • The paper builds systematically on prior work (Refs. 20, 21), reusing the same database of 50 FCE experimental datasets and the same interaction coefficients W_ij, enabling direct comparison with QA results.
  • The choice of a single RY-gate Ansatz is well-motivated for NISQ constraints, reducing gate errors and decoherence.
  • Multiple backends are tested (simulator, ibm_nazca, ibm_brussels), providing some validation.
  • Weaknesses:

  • The number of trials is very small: only 4 trials for ibm_nazca and 5 for ibm_brussels, compared to 10 for QA and 20 for the simulator. This severely limits statistical reliability, and the error bars in Figures 3 and 5 are correspondingly large.
  • The comparison with QA is not entirely fair — the QA results come from a prior publication with potentially different optimization conditions. A contemporaneous, controlled comparison would be more rigorous.
  • The hyperparameter selection (A=1000, B=7) appears to have been done through pre-testing but no systematic sensitivity analysis is provided.
  • The paper acknowledges that ibm_nazca experiments were cut short due to queue times, which introduces inconsistency in experimental conditions.
  • The claim of "lower residual energies" for NISQ devices is made relative to QA, but the reference solution (SA on CPU) consistently outperforms all quantum approaches, undermining claims of quantum utility.
  • Potential Impact

    The practical impact of this work is modest for several reasons:

    1. Problem scale is small: The largest problem uses 90 logical qubits (N=10, Z=9), which is well within the range of classical solvers. SA consistently finds better solutions, and the paper acknowledges this.

    2. No quantum advantage demonstrated: The paper does not claim quantum advantage over classical methods — rather, it compares two quantum approaches. The framing as "utility of NISQ devices" is somewhat misleading since classical SA outperforms both quantum approaches.

    3. Niche application domain: While FCE for atomic junction fabrication is scientifically interesting, the optimization problem itself is a standard combinatorial optimization that could be solved classically. The quantum computing aspect adds complexity without clear benefit.

    4. Incremental over prior work: This is an incremental extension of Yoneda et al. (Ref. 21), which demonstrated QA for the same problem. The contribution is essentially showing that a different quantum computing paradigm (gate-based) can match or exceed QA performance, primarily because it avoids minor embedding overhead.

    The paper could influence the growing body of work exploring NISQ applications in materials science and nanofabrication, serving as a benchmark comparison between gate-based and annealing quantum approaches for real-world optimization problems.

    Timeliness & Relevance

    The paper addresses a timely question — the practical utility of NISQ devices for real-world problems — which is a central concern in the quantum computing community. The comparison between gate-based and annealing paradigms is relevant as both approaches compete for practical applications. However, the field is increasingly skeptical about NISQ utility claims for optimization problems, given the strong performance of classical heuristics, and this paper inadvertently reinforces that skepticism since SA dominates.

    Strengths & Limitations

    Key Strengths:

  • Addresses a real experimental problem rather than a synthetic benchmark
  • Systematic comparison across multiple quantum hardware platforms
  • Clear explanation of the problem formulation and cost function
  • Practical insight that gate-based devices benefit from avoiding minor embedding
  • Notable Limitations:

  • Very limited trial counts on real hardware (4-5 trials), weakening statistical conclusions
  • Classical SA consistently outperforms all quantum methods, undermining the "utility" framing
  • No error mitigation techniques applied (mthree mentioned only as future work)
  • The W_ij matrix is borrowed from prior work rather than generated anew
  • Problem sizes are trivially solvable classically
  • The paper doesn't discuss wall-clock time comparisons, which would be important for practical utility
  • The VQE with 1000 NFT iterations on a quantum computer is significantly slower than SA for these problem sizes
  • No discussion of the cost (financial, time, energy) of running on quantum hardware versus classical alternatives
  • Overall Assessment

    This paper represents an incremental contribution to the niche intersection of quantum computing and nanofabrication. While it provides a useful data point comparing gate-based and annealing quantum approaches on a real-world problem, it does not demonstrate practical quantum advantage. The limited experimental trials, small problem scales, and superior classical performance significantly constrain its impact. The work is competently executed within its scope but unlikely to broadly influence either the quantum computing or nanofabrication communities.

    Rating:3.5/ 10
    Significance 3Rigor 4Novelty 3.5Clarity 5.5

    Generated Apr 15, 2026

    Comparison History (34)

    vs. An Autonomous Topological Pump
    claude-opus-4.65/11/2026

    Paper 1 proposes a fundamentally novel concept—an autonomous Thouless pump where topological transport occurs without external control—bridging topological physics and quantum thermodynamics/motors. This opens new theoretical directions with broad implications across condensed matter and quantum technology. Paper 2 applies existing NISQ optimization techniques to a specific fabrication problem (Au atomic junctions), representing an incremental application of known methods. While practically useful, it lacks the conceptual novelty and breadth of impact of Paper 1, which introduces a new paradigm for topologically protected autonomous quantum devices.

    vs. Demonstrating Record Fidelity for the Quantum Fourier Transform
    claude-opus-4.64/15/2026

    Paper 2 demonstrates a record-breaking quantum Fourier transform implementation on actual quantum hardware (IBM Heron r3) with 50 qubits, achieving super-exponential speedup over previous methods via the Parity Architecture. The QFT is a fundamental subroutine in many quantum algorithms (e.g., Shor's algorithm), making this broadly impactful. The methodological advance—demonstrating a new architecture that dramatically improves scaling—has significant implications for practical quantum computing. Paper 1 addresses a niche application (Au atomic junction fabrication) with incremental improvements over existing quantum annealing approaches.

    vs. Complementary Quantum Time Distributions from a Single Operational Protocol
    gpt-5.24/15/2026

    Paper 1 is more conceptually novel, introducing a unified operational protocol that yields two inequivalent, complementary quantum time distributions with a concrete reinterpretation of the Hartman effect. This addresses a long-standing foundational problem (time in quantum mechanics) and may influence quantum measurement theory, tunneling/transport, and time observables broadly. Paper 2 is timely and application-driven, but its impact is likely narrower and contingent on near-term NISQ performance; it mainly benchmarks optimization quality versus a D-Wave approach without clear methodological breakthroughs in fabrication or quantum algorithms.

    vs. Noise-Robust Ultrafast Entanglement Generation in Rydberg Atoms via Quantum Optimal Control
    gpt-5.24/15/2026

    Paper 2 has higher likely impact: it addresses a central bottleneck for neutral-atom quantum computing—high-fidelity, noise-robust ultrafast entangling gates—using rigorous modeling (multiple realistic noise spectra, Monte Carlo ensembles) and a concrete control-design method (QOCT/D-MORPH) that yields actionable pulse features and quantitative noise thresholds. Its results are timely and broadly relevant to quantum information, AMO physics, and control engineering. Paper 1 is interesting but narrower (a specific fabrication optimization use-case) and its practical advantage over classical optimization baselines and end-to-end experimental validation are less clear from the abstract.

    vs. Testing the 3D QRNG by Undoing
    gpt-5.24/15/2026

    Paper 1 offers a broadly applicable, conceptually novel certification method for photonic 3D QRNGs by “undoing” the implemented unitary, directly addressing correctness, noise, loss, and systematic errors, and linking to foundational constraints (WAY theorem). This strengthens trust in quantum randomness—high real-world relevance for cryptography and secure systems—and could generalize to other photonic quantum devices, boosting cross-field impact. Paper 2 is timely but more incremental: applying NISQ optimization to a specific fabrication-control task, with impact likely limited by current NISQ scalability and practicality versus classical or specialized optimization.

    vs. Entanglement concentration via measurement:- role of imaginarity
    claude-opus-4.64/15/2026

    Paper 2 addresses fundamental questions in quantum information theory—the operational role of imaginarity as a resource—with concrete, quantifiable advantages (22.7% reduction in bond occupation probability for quantum network percolation). It connects resource theory formalism to practical protocols (entanglement concentration, swapping, network percolation), resolves open problems in the literature, and has broader theoretical implications across quantum foundations and quantum networking. Paper 1 presents an incremental application of NISQ devices to a niche fabrication problem with limited generalizability.

    vs. Quantifying magic via quantum $(α,β)$ Jensen-Shannon divergence
    gemini-34/15/2026

    Paper 1 presents a practical, cross-disciplinary application of near-term quantum hardware to solve real-world nanofabrication problems. By demonstrating the utility of NISQ devices for optimizing experimental parameters and comparing them favorably to quantum annealers, it bridges quantum computing with materials science. In contrast, Paper 2 offers a valuable but strictly theoretical contribution to quantum resource theory. Paper 1's tangible experimental relevance and immediate applicability give it a broader potential scientific impact.

    vs. Emission and Absorption of Microwave Photons in Orthogonal Temporal Modes across a 30-Meter Two-Node Network
    gemini-34/15/2026

    Paper 1 presents a foundational experimental breakthrough in quantum networking by demonstrating the transfer of microwave photons in orthogonal temporal modes over a 30-meter cryogenic link. This significantly advances the physical layer of quantum communication and distributed quantum computing. In contrast, Paper 2 offers a narrower, application-specific use of existing NISQ devices for parameter optimization in materials fabrication. Paper 1's contribution to enabling a new photonic degree of freedom in waveguide quantum electrodynamics promises a much broader and more fundamental scientific impact.

    vs. Permutationally symmetric molecular aggregates
    claude-opus-4.64/15/2026

    Paper 2 makes a fundamental theoretical contribution by rigorously connecting classical optics approximations (DDA/CES/CPA) to quantum mechanical treatments of molecular aggregates, identifying exact validity limits and systematic 1/N corrections. This bridges quantum optics, molecular physics, and polaritonics, with broad implications for understanding quantum optical effects in molecular systems. Paper 1 presents an incremental application of NISQ devices to a specific fabrication problem, with narrower scope and less generalizable insights. Paper 2's theoretical framework has greater potential to influence multiple research communities.

    vs. Deterministic quantum master equation for non-Markovian signal processing
    gpt-5.24/15/2026

    Paper 1 likely has higher scientific impact: it introduces a broadly applicable deterministic master-equation framework for measurement-based feedback with arbitrary non-Markovian (memory) signal processing, advancing foundational open-quantum-systems theory and providing a tool usable across quantum control, metrology, and quantum technologies. Its novelty and potential for cross-field reuse are high, and it addresses timely needs to model realistic frequency-dependent feedback. Paper 2 is application-focused and timely, but its impact is narrower (specific fabrication optimization) and depends strongly on near-term NISQ performance and benchmarking details.

    vs. Quantum chaos in many-body systems of indistinguishable particles
    gemini-34/15/2026

    Paper 2 offers a broad, unifying theoretical framework for quantum chaos and many-body interference, touching on highly influential topics like out-of-time-order correlators and random-matrix theory. This foundational approach drives significant advancements across multiple physics domains, including condensed matter, high-energy physics, and quantum information. In contrast, Paper 1 presents a valuable but highly specific application benchmarking NISQ devices for atomic junction fabrication. While Paper 1 demonstrates near-term technological utility, Paper 2's fundamental theoretical contributions provide a wider breadth of impact and long-term scientific relevance.

    vs. Scattering Faddeev calculations in the double continuum
    gpt-5.24/15/2026

    Paper 2 likely has higher scientific impact: it targets a timely, fast-moving area (NISQ-era quantum computing) and links it to an applied, automation-relevant materials/nanofabrication problem (autonomous optimization of FCE parameters for Au junctions). If validated, it could influence experimental workflows, quantum optimization benchmarking, and hybrid quantum-classical control across disciplines. Paper 1 is methodologically rigorous and valuable for few-body nuclear physics, but its scope is narrower and more incremental within a mature field, with limited cross-field application.

    vs. Floquet Many-Body Cages
    claude-opus-4.64/15/2026

    Paper 2 introduces a fundamentally new theoretical framework—Floquet many-body cages—that connects several frontier topics: nonergodic quantum dynamics, Floquet engineering, time crystals, and topological phases. It provides general construction principles applicable to broad classes of quantum circuits and constrained models (e.g., Rydberg arrays), with deep implications for nonequilibrium quantum physics. Paper 1 is a more incremental, application-specific study comparing NISQ device performance for a niche fabrication optimization problem, with narrower scope and limited broader impact.

    vs. Superradiant phase transition in cavity magnonics via Floquet engineering
    gemini-34/15/2026

    Paper 2 demonstrates higher potential scientific impact due to its interdisciplinary approach and immediate real-world applications. By bridging quantum computing (NISQ devices) and materials science (fabricating atomic junctions), it provides practical evidence of quantum utility for complex optimization problems. Furthermore, its benchmarking against D-Wave quantum annealers offers highly relevant and timely insights for the quantum computing community. In contrast, Paper 1, while methodologically rigorous and novel in fundamental physics, is confined to a narrower theoretical subfield (cavity magnonics) and lacks the broader cross-disciplinary applicability of Paper 2.

    vs. Notes on some inequalities, resulting uncertainty relations and correlations. 1. General mathematical formalism
    gpt-5.24/15/2026

    Paper 1 is more likely to have higher impact due to its timely, application-driven contribution at the intersection of NISQ quantum optimization and autonomous experimental control for nanofabrication (Au atomic junctions). Demonstrating improved solution quality vs. a D-Wave annealer suggests practical performance relevance and potential transfer to broader experimental-parameter optimization problems. Paper 2 provides a rigorous mathematical consolidation/extension of known inequality-based uncertainty relations and correlation-matrix formulations, but such work is typically incremental unless it yields distinctly new, widely adopted bounds or operational advantages.

    vs. Quantum machine learning for the quantum lattice Boltzmann method: Trainability of variational quantum circuits for the nonlinear collision operator across multiple time steps
    claude-opus-4.64/15/2026

    Paper 1 addresses a fundamental challenge in quantum computational fluid dynamics—handling nonlinearity in quantum lattice Boltzmann methods—which is a broadly relevant problem across quantum computing and physics. It presents novel architectures (R1/R2 models) with clear theoretical contributions. Paper 2, while interesting in demonstrating NISQ utility for a specific materials fabrication task, has narrower scope and incremental contribution (comparing gate-based vs. annealing approaches). Paper 1's intersection of QML with CFD has broader potential impact across multiple scientific domains.

    vs. Noise-Enhanced Self-Healing Dynamics in Non-Hermitian Systems
    gpt-5.24/15/2026

    Paper 1 offers a broadly novel and timely contribution to non-Hermitian physics by identifying and analytically explaining noise-enhanced self-healing, including distinct weak/strong-noise mechanisms with finite-time Lyapunov and perturbative theory. This is methodologically rigorous and likely generalizable across wave/edge dynamics platforms (photonics, acoustics, cold atoms), with clear guidance for experiments in realistic noisy settings. Paper 2 is more application-specific and benchmarking-focused (NISQ vs annealer for a particular parameter-optimization task), with impact constrained by rapid hardware churn and unclear demonstration of end-to-end experimental autonomy.

    vs. Noise-enhanced quantum kernels on analog quantum computers
    claude-opus-4.64/15/2026

    Paper 1 introduces novel analog and hybrid quantum kernels with a counterintuitive finding that operational noise can enhance performance, providing both theoretical insight (expressivity/model complexity) and practical applications (non-Markovianity estimation). This has broader impact across quantum computing, machine learning, and open quantum systems. Paper 2 demonstrates a useful but narrower application of NISQ devices for optimizing atomic junction fabrication parameters, comparing gate-based QC to quantum annealing. While practically relevant, it is more incremental and application-specific with less generalizable theoretical contribution.

    vs. The Impact of Qubit Connectivity on Quantum Advantage in Noisy IQP Circuits
    gemini-34/15/2026

    Paper 2 addresses a fundamental and critical challenge in quantum computing: the impact of hardware connectivity and noise on achieving quantum advantage. Its findings broadly influence quantum hardware design, complexity theory, and the evaluation of near-term quantum devices. In contrast, Paper 1 presents a relatively narrow, albeit interesting, application of existing NISQ devices to a specific materials fabrication problem. The foundational nature and broad applicability of Paper 2 across the entire field of quantum information science give it a significantly higher potential for scientific impact.

    vs. Opportunistic QKD: Exploiting Idle Capacity of Classical WDM Systems
    gemini-34/15/2026

    Paper 2 addresses a major bottleneck in quantum technologies—integrating Quantum Key Distribution (QKD) with existing classical infrastructure. By exploiting idle WDM channels, it offers a highly practical, scalable solution with broad implications for cybersecurity and telecommunications. Paper 1 is an interesting application of NISQ devices to a specific nano-fabrication problem, but its impact is narrower compared to the potential global deployment of secure quantum communication networks proposed in Paper 2.