Pith. sign in

REVIEW 2 major objections 5 minor 30 references

Scalable suppression of heating errors in large trapped-ion quantum processors

T0 review · 2 major / 5 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 Optimizing Rabi-frequency waveforms against an analytically derived heating-error bound suppresses heating-induced infidelity in Mølmer–Sørensen gates by up to an order of magnitude in large ion crystals.

desk verdict A genuinely useful pulse-optimization framework for heating errors in multi-mode ion traps, with solid small-N evidence, but the large-N headline claims rest on a proxy whose calibration is not established beyond N=6. read the letter →

arxiv 2507.13457 v1 pith:QFHAL2OP submitted 2025-07-17 quant-ph

Zixuan Huo , Yangchao Shen , Xiao Yuan , Xiao-Ming Zhang This is my paper · ORCID
classification quant-ph
keywords trapped-ionquantumcomputingMølmer–Sørensengatemotionalheatingheating-errorsuppressioncontrolquadraticallyconstrainedquadraticprogrammingphase-spacetrajectoriesRabi-frequencyoptimization
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 claims that the leading obstacle to scaling trapped-ion quantum gates—incoherent motional heating—can be suppressed by optimizing the Rabi-frequency waveforms of a Mølmer–Sørensen gate against an efficiently computable upper bound on heating-induced infidelity. The optimization is recast as a positive-semidefinite quadratically constrained quadratic program by dropping cross terms and flipping the sign of negative eigenvalues of the rotation-angle matrix, so it can be solved with standard techniques instead of exponential fidelity estimation. In numerical simulations for chains of up to 55 ions, the optimized pulses reduce estimated heating errors by up to an order of magnitude compared with conventional pulses designed only for frequency robustness, and they beat pulses that simply minimize laser power. At $N=44$ the estimated error is $1.82\times 10^{-3}$, versus $8.72\times 10^{-3}$ for the conventional baseline and $2.18\times 10^{-2}$ for the minimum-Rabi baseline. A sympathetic reader would take the paper to establish a practical route to heating-robust entangling gates in large ion crystals without polychromatic lasers.

What carries the argument

The load-bearing object is the pair of quadratic forms $\widetilde{H}$ and $\widetilde{M}$ built from the chosen pulse basis: $\widetilde{H}$ encodes the heating-error cost and $\widetilde{M}$ encodes the rotation angle after the 'positive-extraction approximation.' The approximation has two steps: Theorem 1 drops the cross terms $E_{p,q}$ and $E_{q,p}$ by bounding them with the diagonal terms, and Theorem 2 projects the rotation-angle matrix onto the kernel of the displacement and robustness constraints, then flips negative eigenvalues to positive ones, producing a positive-semidefinite $\widetilde{M}$ and reflection matrices $S_p,S_q$ that reconstruct the two ion waveforms from one coefficient vector. This turns an NP-hard non-convex problem into a convex PSD-QCQP solved by Lagrange multipliers, making the heating-error minimization scalable.

What would settle it

Simulate or measure actual heating-induced infidelity for the optimized, conventional, and min-Rabi waveforms on a chain with $N=8$ to $20$ ions: if the ratio of $E$ to the true infidelity departs from its small-$N$ value, or if the optimized pulses no longer beat the conventional baseline, the order-of-magnitude large-$N$ claim fails.

Watch

Extended reading notes

Core claim

The paper establishes that heating-induced infidelity in a Mølmer–Sørensen gate is governed by the phase-space trajectories $\alpha_j^m(t)$ of the motional modes through the bound $E=\sum_{j_1,j_2=p,q}|\sum_m(\Gamma^\uparrow_m+\Gamma^\downarrow_m)\int_0^\tau dt\,\alpha^{m*}_{j_1}(t)\alpha^m_{j_2}(t)|$, and that minimizing this bound over the pulse coefficients is a tractable optimization. The absolute values are removed by a Cauchy–Schwarz and arithmetic–geometric-mean bound showing that cross terms never exceed the diagonal terms, and the non-convex rotation-angle constraint is made positive-semidefinite by projecting onto the kernel of the displacement constraints and replacing negative eigenvalues with their absolute values. The result is a positive-semidefinite QCQP whose solution gives the two Rabi waveforms through reflection matrices. Direct simulation for up to six ions confirms that the optimized pulses give the lowest infidelity at every tested detuning, and for larger systems, where only the bound can be evaluated, the estimated error falls with ion number while both baselines rise.

Load-bearing premise

The large-system claims rest on the assumption that the cost function $E$ tracks the true heating infidelity at ion numbers where the exact fidelity cannot be simulated; Appendix D finds $E$ is consistently about four times the simulated infidelity for $N\le 6$ but does not explain why, so the factor is assumed to stay roughly constant as $N$ grows.

Editorial extensions

If this is right

  • The optimized pulses reduce estimated heating errors by up to an order of magnitude at $N=44$, and unlike both baselines the estimated error decreases as the ion number grows.
  • The framework works with any pulse basis, any detuning, and any ion number, and it does not require polychromatic lasers, so it can be implemented on standard monochromatic amplitude-modulated setups.
  • Because the displacement and detuning-robustness constraints are built in, the pulses remain frequency-robust Mølmer–Sørensen gates while suppressing heating.
  • Waveforms optimized for a nominal heating rate remain effective when the actual rates differ, so rough experimental estimates of $\Gamma^\uparrow$ and $\Gamma^\downarrow$ suffice.
  • The method is compatible with existing error-mitigation techniques for laser phase and frequency noise.

Reading between the lines

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

  • The paper does not claim, but its cost function suggests, that the same optimization could be extended to other entangling operations or to whole gate sequences, since $E$ is computed from phase-space trajectories rather than from a specific pulse ansatz.
  • The eigenvalue-flipping construction that makes $\widetilde{M}$ positive-semidefinite is general: any quadratic constraint of the form $c_p^T M c_q=\Theta$ could likely be handled the same way, which may transfer to other control problems with bilinear constraints.
  • A direct hardware test on a chain with more than six ions would settle whether the near-constant factor between $E$ and simulated infidelity persists; the paper's own Appendix D leaves that unexplained.
  • If the proxy is validated, combining this pulse optimization with recooling schedules, which the paper notes are needed anyway after several gates, should further reduce accumulated heating errors in repeated operations.
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.

Referee Report

2 major / 5 minor

Summary. The manuscript proposes a control-optimization framework for suppressing motional heating errors in Mølmer–Sørensen gates. It uses a phase-space heating cost E (Eq. 1), re-derived in Appendix B, and reduces pulse design to a positive-semidefinite QCQP (Eq. 8) via two approximations: dropping cross terms (Theorem 1) and eigenvalue-flipping to make the rotation constraint positive semidefinite (Theorem 2). With piecewise-constant waveforms, the authors report direct Lindblad infidelities for N≤6 and surrogate Eq. (7) estimates up to N=55, claiming up to an order-of-magnitude improvement over conventional and min-Rabi baselines.

Significance. The work addresses an important scalability bottleneck for trapped-ion processors, and the small-N validation is a genuine strength: the optimized waveforms beat two baselines in direct simulations across detunings, and the cost function is derived rather than merely imported. The framework is flexible with respect to pulse basis and noise parameters and is compatible with existing frequency-robust techniques. The principal weakness is that the headline large-N claim rests on an uncalibrated proxy, so the broader significance depends on an assumption that is not yet tested.

major comments (2)
  1. [Numerical result; Fig. 2(d); Eq. (7); Appendix D] The paper's central claim of "up to an order-of-magnitude reduction in infidelities" for up to 55 qubits is not supported by the data as presented. For N>6, Fig. 2(d) plots the surrogate E from Eq. (7), not a simulated infidelity. Appendix D calibrates Eq. (7) against master-equation infidelity only for N≤6, reporting a factor of about four but noting that the gap narrows as fidelity improves and that "the exact origin of this near-constant factor remains unclear." Since the optimized waveforms have much smaller Eq. (7) values than the baselines, any drift of this conversion factor with N or with error magnitude changes the reported improvement ratio. To make the large-N claim, the paper should either restrict it to the small-N direct simulations or add a calibration of Eq. (7) to master-equation infidelity at intermediate N (e.g., N=8–12 with appropriately truncated phonon spaces) before interpreting Fig. 2(d) as an infidelity reduction.
  2. [Method; Theorem 1; Eqs. (5a)–(7)] The reduction from Eq. (5a) to Eq. (7) discards the cross terms |E_pq| + |E_qp| even though Theorem 1 only bounds them by the diagonal terms; the surrogate is therefore not guaranteed to preserve the ordering of waveforms under the original cost E. This matters because Eq. (7) is both the optimization target and the evaluation metric for N>6. The small-N simulations show indirectly that the approximation is reasonable in that regime, but they do not establish it for larger N. A bound on the error introduced by dropping the cross terms, or a numerical comparison between optimizing Eq. (7) and optimizing the full E for an intermediate N, would strengthen the framework's claim to generality.
minor comments (5)
  1. [Abstract and main text] The abstract contains "due to is incoherence nature" and should read "due to its incoherent nature"; the main text also contains typos such as "our methed" and "choosen."
  2. [Proof of Theorem 1, Eq. (10)] In the proof of Theorem 1, Eq. (10) writes α_j1^m(τ)α_j2^m(τ) inside the t-integral; it should be α_j1^m(x)α_j2^m(x) to match the definition of F_j(x).
  3. [Appendix D] The calibration statement in Appendix D is internally ambiguous: the estimates are said to be "about four times larger" than the infidelities, and then "this factor fluctuates around 0.01." Please clarify whether the ratio or the absolute difference is meant.
  4. [Fig. 2 caption] The Fig. 2 caption says "20(N+1) laser detunings μ evenly distributed among N modes," which is unclear; Appendix D's phrasing "dividing the laser detuning into N+1 intervals, each divided into 20 points" is clearer and should be used consistently.
  5. [Reproducibility] Please add a data and code availability statement, since the numerical claims, especially the large-N estimates, are otherwise hard to reproduce.

Circularity Check

1 steps flagged · score 6.0 of 10

Large-N improvement is reported in the same cost function that is optimized; small-N benchmarks keep the core idea partly independent.

  1. fitted input called prediction [Numerical result section (N > 6 paragraph), Fig. 2(d), and Method Eq. (8)]
    "For larger ion numbers (N >6), numerical calculation of the infidelity becomes challenging due to the exponential increase of the Hilbert space dimension. In such cases, we use the upper bound of infidelity after the positive-extraction approximation given in Eq. (7) as an error estimation. ... At N = 44, our method results in a heating error estimation of only 1.82×10−3, a much smaller value compared to the conventional method (8.72×10−3) and the min-Rabi method (2.18×10−2)."

    Eq. (7) is exactly the diagonal heating cost E_pp + E_qq that the optimization is built to minimize: in Eq. (8), H̃ = S_p^T H(p,p)S_p + S_q^T H(q,q)S_q, so the objective c^T H̃c is the Eq. (7) estimate after the positive-extraction projection. The conventional and min-Rabi baselines are feasible pulses for the same constraints (rotation angle, zero displacement, robustness), so any successful minimization of Eq. (8) must return a lower Eq. (7) value than either baseline. Reporting that minimized value at N = 44 as an 'error estimation' and using it to claim order-of-magnitude infidelity reduction is therefore reporting the optimized objective itself, not an independent prediction of infidelity.

full rationale

The paper re-derives the heating-error bound in Appendix B rather than merely importing it, and its small-N (N ≤ 6) results are genuine master-equation simulations, so the self-citation to Ref. [13] is not load-bearing. The central circularity is confined to the large-N demonstration: for N > 6 the quantity plotted and compared is the same cost function that was minimized, so the reduction in that quantity is forced by the optimization (any feasible baseline must have objective no smaller than the optimum). This makes the headline 'order-of-magnitude reduction in infidelities' for large systems a restatement of the optimization objective unless the Eq. (7)-to-infidelity conversion is independently established at large N, which the paper does not do. Appendix D limits the calibration to N ≤ 6 and reports that the estimate-to-infidelity ratio is only approximately constant. Score 6 reflects partial circularity: the N ≤ 6 benchmarks provide independent content, but the large-N central claim reduces by construction.

Assumptions & free parameters 7 free parameters · 7 assumptions · 0 invented entities

The framework rests on the first-order bound E (Eq. 1), re-derived in Appendix B from the self-cited Ref. [13]; on standard trapped-ion domain assumptions (Markovian quasi-static heating, Magnus-truncated gate unitary, product initial states); and on two tractability approximations introduced here, dropping the objective cross terms (Theorem 1) and flipping the sign of the rotation-angle matrix spectrum (Theorem 2). The free parameters are mostly simulation choices: basis size L = 6N + 20, phonon cutoff N_cut = 10, heating rates 100 phonon/s, gate duration 150 μs, plus undisclosed trap parameters, initial phonon state, and target rotation angle. The most fragile ad hoc element is the extrapolated constancy of the roughly fourfold ratio between E and the simulated infidelity for N > 6.

free parameters (7)
  • basis size L = 6N + 20
    Number of piecewise-constant waveform segments, chosen by hand in the numerical results section; the claim that a few segments suffice depends on this choice.
  • phonon Fock cutoff N_cut = 10
    Hilbert-space truncation in the Lindblad simulations for N ≤ 6; the paper acknowledges it inflates conventional-baseline infidelities, so it directly affects the reported improvement ratios.
  • heating rates Γ↑_m = Γ↓_m = 100 phonon/s
    Noise-level input used for the cost function and for the headline simulations; robustness to misestimated Γ is tested only for a few scaling scenarios.
  • gate duration τ = 150 μs
    Chosen for the simulations; whether the advantage persists for shorter gates is not examined.
  • trap parameters (weak and strong traps) = not disclosed
    Mode frequencies, mode eigenvectors, and Lamb-Dicke parameters determine the matrices in Appendix A and the detuning axis of Fig. 2; without them the numerics are not reproducible.
  • initial phonon state = not disclosed
    The infidelity simulations require the initial phonon temperature, and the bound's derivation requires zero phonon first moments; neither is stated in the main text.
  • target rotation angle Θ_targ = π/4 in Appendix C figure
    The main-text simulations do not state the gate angle; the QCQP constraint and all infidelity values depend on it.
assumptions (7)
  • domain assumption Lindblad master equation with Markov, quasi-static heating rates (Eq. A6)
    Heating noise is modeled as independent phonon excitation and relaxation channels with constant Γ↑, Γ↓ during the gate; the framework is not tested against non-Markovian or time-dependent (colored) noise, listed as future work in the discussion.
  • domain assumption Gate evolution is the two-term Magnus form U(τ) = exp(Σ_j φ_j σ_j + iΘ σ_p σ_q) (Eq. A2)
    Standard MS-gate approximation discarding higher-order Magnus terms and assuming Lamb-Dicke and rotating-wave approximations (Eq. A1); the bound in Eq. (1) inherits these.
  • domain assumption Initial spin-phonon product state with zero phonon first moments ⟨a_m⟩ = 0
    Appendix B uses this to cancel the mixed terms (B18)-(B19); the main text says only that the initial state is separable, so the zero-moment condition (ground or thermal states) is implicit.
  • standard math Cauchy-Schwarz, AM-GM, and Lagrange-multiplier solution of the PSD-QCQP
    Used in the proofs of Theorems 1 and 2 and in solving Eq. (8); standard results, no independent check needed.
  • ad hoc to paper Cross terms E_pq and E_qp are dropped from the objective (Theorem 1)
    The bound (6) guarantees the true objective is within a factor of 2 of E_pp + E_qq, but the optimality gap of the resulting waveform design is unquantified.
  • ad hoc to paper Eigenvalue flipping PMP → \tilde M (Theorem 2) preserves a meaningful optimum
    The construction enlarges the feasible set: any c with c^T \tilde M c = Θ_targ yields a valid gate, but the QCQP optimum is not necessarily the optimum of the original non-convex problem (5).
  • ad hoc to paper The ~4x discrepancy between E and simulated infidelity remains roughly constant for N > 6
    Appendix D establishes the factor only for N ≤ 6 and states its origin is unclear; the large-N claims up to 55 ions extrapolate this factor.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Scalable suppression of heating errors in large trapped-ion quantum processors." pith.science (2026). https://pith.science/paper/QFHAL2OP

@misc{pith2026250713457,
  author       = {Pith},
  title        = {Pith review of: Scalable suppression of heating errors in large trapped-ion quantum processors},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QFHAL2OP}},
  note         = {Machine review of arXiv:2507.13457}
}
read the original abstract

Trapped-ion processors are leading candidates for scalable quantum computation. However, motional heating remains a key obstacle to fault-tolerant operation, especially when system size increases. Heating error is particularly challenging to suppress due to is incoherence nature, and no general methods currently exist for mitigating their impact even in systems with more than two ions. In this work, based on a careful analysis about the dependence of heating-induced infidelity on phase-space trajectories, we present a simple yet comprehensive framework for suppressing heating errors in large trapped-ion quantum processors. Our approach is flexible, allowing various control pulse bases, ion numbers, and noise levels. Our approach is also compatible with existing error-mitigation techniques, including those targeting laser phase and frequency noise. Crucially, it relies on an efficiently computable cost function that avoids the exponential overhead of full fidelity estimation. We perform numerical simulations for systems with up to 55 qubits, demonstrating up to an order-of-magnitude reduction in infidelities. These results offer a practical route toward robust, large-scale quantum computation with trapped ions.

Figures

Figures reproduced from arXiv: 2507.13457 by the authors.

Figure 1
Figure 1. FIG. 1: Framework of heating error suppression. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2: Infidelities, estimations and Rabi [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3: (a)-(f): Trajectory of [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (3 more)
Figure 5
Figure 5. Figure 5: FIG. 5: Improve rate of our method and previous method with 2 ions, excitation and relaxation rate [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: FIG. 6: Infidelity versus laser detuning [PITH_FULL_IMAGE:figures/full_fig_p015_6.png]
Figure 7
Figure 7. Figure 7: shows the 𝐿2-norm of Rabi frequencies for three different methods, similar to [PITH_FULL_IMAGE:figures/full_fig_p016_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 28 canonical work pages

  1. [1]

    Thus, all three constraints are simultaneously satisfied, com- pleting the proof

    The constraint ®𝑐𝑇 𝑝M®𝑐𝑞 = Θtarg is satisfied because ®𝑐𝑇 𝑝M®𝑐𝑞 =®𝑐𝑇PMPF®𝑐=®𝑐𝑇 ˜M®𝑐=Θ targ; 2.®𝑐 𝑝 and®𝑐𝑞 lie in the kernel of both A and Adiff because they are projections via P. Thus, all three constraints are simultaneously satisfied, com- pleting the proof. Acknowledgement We thank Wenhao Zhang for helpful discussions. This work is supported by the Na...

    work page

  2. [2]

    Quantum computing with trapped ions

    Hartmut H ¨affner, Christian F Roos, and Rainer Blatt. Quantum computing with trapped ions. Physics reports, 469(4):155–203, 2008

    work page 2008

  3. [3]

    Scaling the ion trap quantum processor

    Christopher Monroe and Jungsang Kim. Scaling the ion trap quantum processor. Science, 339(6124):1164–1169, 2013

    work page 2013

  4. [4]

    Trapped-ion quantum computing: Progress and challenges

    Colin D Bruzewicz, John Chiaverini, Robert McConnell, and Jeremy M Sage. Trapped-ion quantum computing: Progress and challenges. Applied physics reviews, 6(2), 2019

    2019

  5. [5]

    High-fidelity prepara- tion, gates, memory, and readout of a trapped-ion quantum bit

    TP Harty, DTC Allcock, C J˜ Ballance, L Guidoni, HA Janacek, NM Linke, DN Stacey, and DM Lucas. High-fidelity prepara- tion, gates, memory, and readout of a trapped-ion quantum bit. Physical review letters, 113(22):220501, 2014

    work page 2014

  6. [6]

    A race-track trapped-ion quantum proces- sor

    Steven A Moses, Charles H Baldwin, Michael S Allman, R An- cona, L Ascarrunz, C Barnes, J Bartolotta, B Bjork, P Blan- chard, M Bohn, et al. A race-track trapped-ion quantum proces- sor. Physical Review X, 13(4):041052, 2023

    work page 2023

  7. [7]

    The computa- tional power of random quantum circuits in arbitrary geometries

    Matthew DeCross, Reza Haghshenas, Minzhao Liu, Yuri Alex- eev, Charles H Baldwin, John P Bartolotta, Matthew Bohn, Eli Chertkov, Jonhas Colina, Davide DelVento, et al. The computa- tional power of random quantum circuits in arbitrary geometries. arXiv preprint arXiv:2406.02501, 2024

    arXiv 2024

  8. [8]

    Continuous symmetry breaking in a trapped-ion spin chain

    Lei Feng, Or Katz, Casey Haack, Mohammad Maghrebi, Alexey V Gorshkov, Zhexuan Gong, Marko Cetina, and Christo- pher Monroe. Continuous symmetry breaking in a trapped-ion spin chain. Nature, 623(7988):713–717, 2023

    work page 2023

Show all 30 references
  1. [9]

    Non-abelian topological order and anyons on a trapped-ion pro- cessor

    Mohsin Iqbal, Nathanan Tantivasadakarn, Ruben Verresen, Sara L Campbell, Joan M Dreiling, Caroline Figgatt, John P Gaebler, Jacob Johansen, Michael Mills, Steven A Moses, et al. Non-abelian topological order and anyons on a trapped-ion pro- cessor. Nature, 626(7999):505–511, 2024

    2024

  2. [10]

    Characterizing a non-equilibrium phase transition on a quantum computer

    Eli Chertkov, Zihan Cheng, Andrew C Potter, Sarang Gopalakr- ishnan, Thomas M Gatterman, Justin A Gerber, Kevin Gilmore, Dan Gresh, Alex Hall, Aaron Hankin, et al. Characterizing a non-equilibrium phase transition on a quantum computer. Nature Physics, 19(12):1799–1804, 2023

    2023

  3. [11]

    High-fidelity quantum logic gates using trapped-ion hyperfine qubits.Physicalreviewletters, 117(6):060504, 2016

    Christopher J Ballance, Thomas P Harty, Nobert M Linke, Mar- tin A Sepiol, and David M Lucas. High-fidelity quantum logic gates using trapped-ion hyperfine qubits.Physicalreviewletters, 117(6):060504, 2016

    2016

  4. [12]

    High-fidelity light-shift gate for clock-state qubits

    CH Baldwin, BJ Bjork, M Foss-Feig, JP Gaebler, D Hayes, MG Kokish, C Langer, JA Sedlacek, D Stack, and G Vittorini. High-fidelity light-shift gate for clock-state qubits. Physical Review A, 103(1):012603, 2021

    2021

  5. [13]

    Noise analysis for 9 high-fidelity quantum entangling gates in an anharmonic linear paul trap

    Yukai Wu, Sheng-Tao Wang, and L-M Duan. Noise analysis for 9 high-fidelity quantum entangling gates in an anharmonic linear paul trap. Physical Review A, 97(6):062325, 2018

    2018

  6. [14]

    Scaling of entangling-gate errors in large ion crystals

    Wenhao He, Wenhao Zhang, Xiao Yuan, Yangchao Shen, and Xiao-Ming Zhang. Scaling of entangling-gate errors in large ion crystals. Journal of Physics A: Mathematical and Theoretical, 57(37):375306, 2024

    2024

  7. [15]

    High fidelity quan- tum gates of trapped ions in the presence of motional heating

    Farhang Haddadfarshi and Florian Mintert. High fidelity quan- tum gates of trapped ions in the presence of motional heating. New Journal of Physics, 18(12):123007, dec 2016

    2016

  8. [16]

    Robust entanglement gates for trapped-ion qubits

    Yotam Shapira, Ravid Shaniv, Tom Manovitz, Nitzan Akerman, and Roee Ozeri. Robust entanglement gates for trapped-ion qubits. Physical review letters, 121(18):180502, 2018

    2018

  9. [17]

    Resilient entangling gates for trapped ions

    Anna E Webb, Simon C Webster, S Collingbourne, David Bre- taud, Adam M Lawrence, Sebastian Weidt, Florian Mintert, and Winfried K Hensinger. Resilient entangling gates for trapped ions. Physical review letters, 121(18):180501, 2018

    2018

  10. [18]

    Multiparticle entangle- ment of hot trapped ions

    Klaus Mølmer and Anders Sørensen. Multiparticle entangle- ment of hot trapped ions. Physical Review Letters, 82(9):1835, 1999

    1999

  11. [19]

    Quantum computation with ions in thermal motion.Physicalreviewletters, 82(9):1971, 1999

    Anders Sørensen and Klaus Mølmer. Quantum computation with ions in thermal motion.Physicalreviewletters, 82(9):1971, 1999

    1971

  12. [20]

    High-fidelity universal gate set for 9Be+ ion qubits

    John P Gaebler, Ting Rei Tan, Yiheng Lin, Y Wan, Ryan Bowler, Adam C Keith, Scott Glancy, Kevin Coakley, Emanuel Knill, Dietrich Leibfried, et al. High-fidelity universal gate set for 9Be+ ion qubits. Physical review letters, 117(6):060505, 2016

    2016

  13. [21]

    Robust 2-qubit gates in a linear ion crystal using a frequency-modulated driving force

    Pak Hong Leung, Kevin A Landsman, Caroline Figgatt, Norbert M Linke, Christopher Monroe, and Kenneth R Brown. Robust 2-qubit gates in a linear ion crystal using a frequency-modulated driving force. Physical review letters, 120(2):020501, 2018

    2018

  14. [22]

    Ion-trap measure- ments of electric-field noise near surfaces

    M Brownnutt, M Kumph, P Rabl, and R Blatt. Ion-trap measure- ments of electric-field noise near surfaces. Reviews of modern Physics, 87(4):1419, 2015

    2015

  15. [23]

    Measurement of ion motional heating rates over a range of trap frequencies and temperatures

    CD Bruzewicz, JM Sage, and J Chiaverini. Measurement of ion motional heating rates over a range of trap frequencies and temperatures. Physical Review A, 91(4):041402, 2015

    2015

  16. [24]

    Second order gradient ascent pulse engineering

    Pierre de Fouquieres, Sophie G Schirmer, Steffen J Glaser, and Ilya Kuprov. Second order gradient ascent pulse engineering. Journal of Magnetic Resonance, 212(2):412–417, 2011

    2011

  17. [25]

    When does reinforcement learning stand out in quantum control? a comparative study on state preparation

    Xiao-Ming Zhang, Zezhu Wei, Raza Asad, Xu-Chen Yang, and Xin Wang. When does reinforcement learning stand out in quantum control? a comparative study on state preparation. npj Quantum Information, 5(1):85, 2019

    2019

  18. [26]

    Nonadiabatic holonomic quantum computation in decoherence-free subspaces with trapped ions

    Zhen-Tao Liang, Yan-Xiong Du, Wei Huang, Zheng-Yuan Xue, and Hui Yan. Nonadiabatic holonomic quantum computation in decoherence-free subspaces with trapped ions. Physical Review A, 89(6):062312, 2014

    2014

  19. [27]

    Experimental realization of nonadiabatic holonomic single- qubit quantum gates with two dark paths in a trapped ion

    Ming-Zhong Ai, Sai Li, Ran He, Zheng-Yuan Xue, Jin-Ming Cui, Yun-Feng Huang, Chuan-Feng Li, and Guang-Can Guo. Experimental realization of nonadiabatic holonomic single- qubit quantum gates with two dark paths in a trapped ion. Fundamental Research, 2(5):661–666, 2022

    2022

  20. [28]

    Robust mølmer-sørensen gate against symmetric and asymmetric er- rors

    Wenhao Zhang, Gaoxiang Tang, Kecheng Liu, Xiao Yuan, Yangchao Shen, Yukai Wu, and Xiao-Ming Zhang. Robust mølmer-sørensen gate against symmetric and asymmetric er- rors. arXiv preprint arXiv:2501.02847, 2025

    2025 arXiv

  21. [29]

    Pre- cision measurements in ion traps using slowly moving standing waves

    A Walther, U Poschinger, K Singer, and F Schmidt-Kaler. Pre- cision measurements in ion traps using slowly moving standing waves. Applied Physics B, 107:1061–1067, 2012

    2012

  22. [30]

    These derivations clarify the assumptions and computational structures used throughout our method

    This value might subject to numerical error due to the high average phonon excitation 10 Appendix A: definitions and expressions In this appendix, we provide the explicit mathematical expressions underlying the gate construction, the heating error model, and the optimization f...

Pith tools

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