Pith. sign in

REVIEW 3 major objections 5 minor 61 references

An efficient non-unitary loading scheme plus multigrid warm starts lets variational quantum linear solvers handle Carleman-linearized 1D Burgers and transpile circuits with up to 2^80 space-time points.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · grok-4.5

2026-07-13 01:16 UTC pith:ERKGHSFV

load-bearing objection Real engineering progress on LCNU loading + multigrid-VQLS for Carleman-Burgers, with honest hardware numbers; the 2^80 feasibility claim is the soft spot because iteration counts are extrapolated from tiny systems. the 3 major comments →

arxiv 2607.08976 v1 pith:ERKGHSFV submitted 2026-07-09 quant-ph physics.flu-dyn

A Scalable Approach to Solve the Carleman Linearized Burgers' Equation on a Quantum Computer

Reuben Demirdjian , Yvan Quinn , Vincent P. Su , Hrant Gharibyan , Hayk Tepanyan This is my paper
classification quant-ph physics.flu-dyn
keywords Carleman linearizationBurgers equationlinear combination of non-unitariesvariational quantum linear solvermultigrid warm startquantum fluid dynamicsbarren plateau
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

Nonlinear fluid equations are hard for quantum computers because the machines are linear. This paper shows how to turn the one-dimensional Burgers equation into a linear system via Carleman linearization, load that system with a fixed number of non-unitary terms that does not grow with grid size, and solve it with a variational quantum linear solver. Barren plateaus are mitigated by solving a sequence of coarser-to-finer space-time grids, using each stage’s optimum as a warm start for the next. On a simulator the multigrid path recovers wave damping and nonlinear advection that a random cold start misses; single cost-function evaluations also run on real IBM processors. Transpiling the same circuits out to combined discretizations of order 2^80 yields polylogarithmic CNOT depth, which the authors argue makes a future quantum advantage feasible once decoherence improves.

Core claim

The Carleman-linearized, zero-padded 1D Burgers matrix admits an LCNU decomposition whose term count is exactly ½(9α² + α + 8) and therefore independent of the number of spatial and temporal points; when this loading is paired with spatial-temporal multigrid warm starts inside VQLS, the resulting circuits can be solved accurately for small systems and still transpile with polylogarithmic two-qubit depth at 2^80 grid points.

What carries the argument

Linear combination of non-unitaries (LCNU) embedding of the zero-padded Carleman matrix together with a multigrid stage ladder that doubles spatial or temporal resolution while interpolating variational parameters via Hadamard gates on newly added register qubits.

Load-bearing premise

The multigrid warm-start procedure that tames barren plateaus on tiny grids will keep working once the linear system itself becomes exponentially large.

What would settle it

Run the identical multigrid VQLS pipeline on a noise-free simulator (or sufficiently coherent hardware) for a sequence of grids larger than 16×16 and check whether the final-stage cost still reaches ~10⁻³ and the recovered velocity field continues to match a classical reference within a fixed tolerance; failure at the next power-of-two size would falsify the claimed scalability.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Any PDE whose Carleman form admits an LCNU decomposition of size independent of grid resolution can reuse the same loading and multigrid solver stack.
  • Resource estimates for fluid problems become dominated by CNOT depth and shot allocation rather than by classical matrix assembly.
  • Hardware roadmaps can treat ~100-qubit devices with polylog depth as sufficient to enter the regime beyond classical 3-D turbulence grids once preconditioning and readout are solved.
  • Shot budgets can be allocated proportionally to coefficient magnitudes, cutting sample waste on near-zero LCNU terms.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Because the term count grows only with Carleman order α, the dominant remaining cost is likely matrix conditioning rather than circuit width; preconditioning therefore becomes the next bottleneck for any practical advantage.
  • The same Hadamard-interpolation warm-start idea could transfer to other hierarchical quantum algorithms (e.g., multilevel quantum Monte Carlo or nested VQE) that suffer barren plateaus.
  • If larger α is required for high-Reynolds accuracy, the exponential growth of system size with α may erase the polylog grid advantage unless a smarter truncation is found.
  • Selective readout of only the first-order spatial slice already reduces the output problem; pairing it with classical post-processing of low-order moments could make end-to-end fluid diagnostics feasible before full state tomography.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The manuscript develops an end-to-end pipeline for the Carleman-linearized 1D Burgers’ equation: an improved linear combination of non-unitaries (LCNU) loading whose term count Ns = ½(9α² + α + 8) is independent of spatial/temporal resolution, a spatial–temporal multigrid warm-start for VQLS that mitigates barren plateaus on small instances, a variance-minimizing shot-allocation scheme, and resource estimates obtained by transpiling the most expensive circuits onto IBM Heron/Nighthawk topologies up to combined discretizations of 2^80. Simulator results (Figs. 11–12) show multigrid outperforming cold starts; a single VQLS cost-function iteration is executed on real hardware (Figs. 15–16); wall-clock estimates via Eq. (41) are used to argue that quantum advantage is feasible on future devices.

Significance. The work supplies concrete, reusable technical pieces for quantum nonlinear PDE solvers: an LCNU decomposition whose gate depth remains polylogarithmic after transpilation (Fig. 17a), an explicit multigrid interpolation rule adapted to the Carleman zero-padded registers (Alg. 3, Figs. 7–8), and a shot-allocation formula derived from the cost-function variance (Eqs. 33–39). These contributions are checked against classical linear solvers and an ideal state-vector simulator, and the hardware runs and large-scale transpilation are reproducible on public IBM backends. If the multigrid iteration counts remain modest at exponential scale, the pipeline would constitute a credible route toward quantum advantage for a paradigmatic nonlinear fluid equation; even without that, the loading and multigrid constructions are of independent interest to the quantum linear-systems community.

major comments (3)
  1. Abstract and §5.3 convert the polylog CNOT depth of Fig. 17a into an O(10²)-hour wall-clock claim via Eq. (41), which hard-codes N_iter = 25 000 per multigrid stage taken from the (16,16) simulator run of Fig. 11 and N_multi = 2(q−2)+1. The authors themselves list as open caveat 3 in §6 that multigrid has not been shown to suppress barren plateaus once the Hilbert-space dimension is exponential. Because the feasibility-of-advantage headline rests directly on this fixed-iteration assumption, either a stronger scaling argument (or empirical bound) for N_iter, or a clearly softened claim that removes the quantitative hour estimate, is required.
  2. All numerical demonstrations (§5.1–5.2) and the resource model use Carleman truncation α = 2, the minimal order that retains a quadratic nonlinearity. Section 6 correctly notes that larger α may be needed for accuracy and that system size grows exponentially with α (via n_x^α). The abstract and contribution list nevertheless present the pipeline as ready for quantum advantage without quantifying how Ns (Eq. 23) and circuit width scale for α > 2, nor how the multigrid warm-start (Figs. 7–8) generalizes when the number of spatial registers increases. A short scaling table or explicit statement that the present results are restricted to α = 2 would make the central claim precise.
  3. Section 6 acknowledges that the Carleman matrix is poorly conditioned and that preconditioning will be required for larger systems, yet no condition-number estimates, even for the small instances of §5.1, are reported, and the resource model of Eq. (41) omits any preconditioning overhead. Because VQLS convergence and the validity of the fixed N_iter budget both depend on conditioning, at least a classical condition-number plot versus (n_x, n_t, α) for the matrices actually solved would strengthen the feasibility argument.
minor comments (5)
  1. Typographical errors: “combintaiton” (abstract), “Algoirthm” (twice in §4.4), “discretizaitons” (§5), “lineraization” (§6).
  2. Figure 6 caption states “16 variational parameters for the 4-qubit case”; the circuit diagram shows four qubits but the later experiments use up to 9 + 2 ancilla qubits—clarify how the layer count and parameter count scale with register size.
  3. Equation (23) for Ns is derived after cancelling redundancies; a short intermediate count (before and after cancellation) would help readers verify the formula.
  4. The shot-allocation derivation assumes real coefficients and then briefly notes the complex extension; stating explicitly whether all c_l arising from the Burgers LCNU are real would remove ambiguity for implementers.
  5. References [14] and [33] are the authors’ own prior LCNU papers; a sentence distinguishing what is new here (Ns independent of n_t via Eq. 14, multigrid for Carleman registers, shot allocation, hardware runs) from what is reused would improve novelty clarity.

Circularity Check

1 steps flagged

No load-bearing circularity; LCNU decompositions and multigrid are first-principles or empirically validated against classical solvers, with only non-load-bearing self-cites for reusable embedding lemmas.

specific steps
  1. self citation load bearing [Sec. 3.1–3.2 and citations [14, 31, 33]]
    "Using this LCNU approach, an efficient data loading method for the Carleman linearized 1D Burgers’ equation was developed in [14]. … As shown in [14, 31, 33], the L(e) matrix can be decomposed into a polylogarithmic number of terms"

    The base LCNU embedding framework and zero-padding construction are taken from the authors’ own prior works. However the present paper re-derives the concrete circuits, improves the term count of L1 and F1 to be independent of nx/nt, and supplies the full Ns formula; the self-citations therefore supply reusable lemmas rather than the target numerical or feasibility claims, rendering the circularity minor and non-load-bearing.

full rationale

The paper's derivation chain is self-contained. The LCNU decomposition of the zero-padded Carleman-Burgers matrix (Sec. 3.2, Eqs. 14–23) is obtained by direct algebraic expansion of F1, F2, incrementers and commutation matrices, yielding the explicit Ns = ½(9α² + α + 8) that is independent of nx, nt; the unitary embeddings (Algs. 1–2, Figs. 2–4) follow from the same linear-algebra conditions. Multigrid warm-starts (Alg. 3, Figs. 7–10) are a constructive interpolation procedure whose benefit is measured by direct comparison of cost-function trajectories and solution profiles against both a classical linear solver and a cold-start VQLS baseline (Figs. 11–12). Shot allocation (Eqs. 33–39) is an analytic Lagrange-multiplier minimization of cost-function variance. Resource estimates (Eq. 41, Fig. 17) hard-code N_iter from the small-system run of Fig. 11, but the paper itself lists this as an open caveat (Sec. 6, item 3) rather than a derived prediction. Self-citations to the authors’ prior LCNU papers supply reusable embedding lemmas that are re-derived and improved here; they do not force the numerical claims, which are checked against external classical and simulator benchmarks. No step reduces a claimed prediction to a fitted input or self-definition by construction.

Axiom & Free-Parameter Ledger

4 free parameters · 3 axioms · 0 invented entities

The central scalability claim rests on a handful of modeling choices (truncation order, Reynolds number, ansatz depth, fixed iteration budget) and on the unproven extrapolation that multigrid continues to work at exponential size. No new physical entities are postulated; the free parameters are ordinary numerical hyperparameters.

free parameters (4)
  • Carleman truncation order α = 2
    Fixed at the minimal value α=2 throughout; higher α is acknowledged to be necessary for accuracy but is never varied.
  • Reynolds number Re = 5
    Set by hand to Re=5 to keep the flow laminar and dynamically similar across multigrid stages.
  • VQLS iteration budget Niter = 25000
    Fixed at 25 000 per multigrid stage for the resource estimate; chosen by inspecting the small-system cost curves rather than derived.
  • Ansatz layer count / parameter count
    Hardware-efficient RY–RZ–CZ brick-wall depth is chosen empirically; no systematic expressivity analysis is given.
axioms (3)
  • domain assumption Carleman linearization truncated at finite α yields a useful approximation to the nonlinear Burgers dynamics for the chosen laminar regime.
    Invoked throughout §2; the paper itself notes that for higher Re the truncation can diverge (§6).
  • standard math The local VQLS cost (Eq. 25) has a global minimum at the true solution of the linear system.
    Taken from Bravo-Prieto et al. (2023) and used without re-derivation.
  • ad hoc to paper Multigrid warm starts remain effective once the Hilbert-space dimension becomes exponential.
    Stated as an open caveat in §6; only verified up to 16×16 grids.

pith-pipeline@v1.1.0-grok45 · 31760 in / 2728 out tokens · 34039 ms · 2026-07-13T01:16:41.343641+00:00 · methodology

0 comments
read the original abstract

Efficiently solving nonlinear ordinary and partial differential equations using a quantum computer is a major challenge due its inherent linearity. To circumvent this challenge, the Carleman linearization method has been proposed to transform a nonlinear ordinary differential equation into a linear system of equations, the primary advantage being that existing quantum linear systems algorithms may then be applied to obtain a solution. However, this methodology also brings forth several major challenges that must be addressed to attain a quantum advantage. Herein, we address several of these challenges enabling us to solve the Carleman linearized one-dimensional Burgers' equation on real and simulated quantum hardware. All simulations were performed on BlueQubit's platform allowing for quantum circuits to be run on GPU or QPU's seamlessly. We first demonstrate that the Carleman linearized Burgers' equation can be efficiently loaded onto a quantum computer using the linear combination of non-unitaries method, an alternative to the linear combintaiton of unitaries approach. Once loaded, the linear system is then solved using the variational quantum linear solver. Since a naive implementation of this solver is hindered by the barren plateau phenomenon, we introduce a multigridding method to solve the problem in a series of stages with the solution of the previous stage acting as a warm start for the next stage. This approach is found to significantly improve the accuracy of the solution compared with a naive cold start. Finally, circuits with a combined number of spatial and temporal discretization points totaling up to $2^{80} \approx 10^{24}$ are transpiled onto real quantum hardware demonstrating that the proposed methodology could feasibly produce a quantum advantage on future hardware.

Figures

Figures reproduced from arXiv: 2607.08976 by Hayk Tepanyan, Hrant Gharibyan, Reuben Demirdjian, Vincent P. Su, Yvan Quinn.

Figure 1
Figure 1. Figure 1: The circuit from Example 3.1. The |0⟩a wire is a single ancillary qubit required for the unitary embedding and the vertical dashed line separates the U1 from the U2 components. The equations below the circuits are: the non-unitary term to embed (L), its completion (L) which is necessary to construct U2 = I ⊗L (see Algorithm 2), and LLT which is used to construct U1 (see Algorithm 1). Circuits drawn using [… view at source ↗
Figure 2
Figure 2. Figure 2: Same as Figure [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Same as Figure [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Same as Figure [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Circuits for evaluating the real part of the local cost function [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Ansatz used for both the simulator and real IBM hardware training. Here, [PITH_FULL_IMAGE:figures/full_fig_p012_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: (left) The ansatz circuit at the ith multigrid stage with, from top to bottom, α spatial registers each with log n (i) x qubits, the zero-padded register with log α qubits, and the temporal register with log n (i) t qubits where n (i) x and n (i) t are the number of spatial and temporal points, respectively, at the ith stage (not to the ith power). Here, ⃗θ (i) are the starting variational parameters at th… view at source ↗
Figure 8
Figure 8. Figure 8: (left) Exactly as in Figure [PITH_FULL_IMAGE:figures/full_fig_p015_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: A warm start example for an increase in the temporal points from stages [PITH_FULL_IMAGE:figures/full_fig_p016_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: The same as in Figure [PITH_FULL_IMAGE:figures/full_fig_p017_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Cost function evaluations from (25) as a function of optimizer iterations for each of the four multigridding stages (black) and the single naive stage (blue). The negative values on the x-axis indicate iterations prior to the final multigridding stage (stage 4). The naive approach is chosen to start at iteration zero to facilitate comparison with the final multigridding stage. 5.2 Real Hardware To benchma… view at source ↗
Figure 12
Figure 12. Figure 12: The flow velocity u(x) as a function of spatial domain x ∈ [0, 2π) at each time step (color) for the problem setup described in Section 5 using the (a) classical linear system solver numpy.linalg.solve in Python, (b) naive approach whereby the VQLS variational parameters are initialized randomly, and (c-f) the four multigridding (MG) stages where the VQLS variational parameters are initialized using the o… view at source ↗
Figure 13
Figure 13. Figure 13: The two-qubit gate depths of the 2760 transpiled circuits for stage 2 of the multigridding approach [PITH_FULL_IMAGE:figures/full_fig_p021_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Distribution of shots (left y-axis) for the 2760 circuits using [PITH_FULL_IMAGE:figures/full_fig_p022_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: The results for the simulated and real hardware runs in [PITH_FULL_IMAGE:figures/full_fig_p023_15.png] view at source ↗
Figure 16
Figure 16. Figure 16: The results from the Hadamard test for each of the 2760 circuits obtained from real IBM [PITH_FULL_IMAGE:figures/full_fig_p024_16.png] view at source ↗
Figure 17
Figure 17. Figure 17: (a) The two-qubit gate count, and (b) the total executions times using [PITH_FULL_IMAGE:figures/full_fig_p026_17.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

61 extracted references · 6 canonical work pages

  1. [1]

    Quantum computing for nonlinear differential equations and turbulence

    Felix Tennie et al. “Quantum computing for nonlinear differential equations and turbulence.” In:Nature Reviews Physics7.4 (2025), pp. 220–230

    2025

  2. [2]

    Computational fluid dynamics on quantum computers

    Madhava Syamlal et al. “Computational fluid dynamics on quantum computers.” In:AIAA AVIATION FORUM AND ASCEND 2024. 2024, p. 3534

    2024

  3. [3]

    Quantum Computers for Weather and Climate Prediction: The Good, the Bad, and the Noisy

    F. Tennie and T. N. Palmer. “Quantum Computers for Weather and Climate Prediction: The Good, the Bad, and the Noisy.” In:Bulletin of the American Meteorological Society104.2 (2023), E488–E500.doi: 10.1175/BAMS-D-22-0031.1 .url: https://journals.ametsoc.org/view/journals/bams/104/2/ BAMS-D-22-0031.1.xml

    doi:10.1175/bams-d-22-0031.1 2023

  4. [4]

    Efficient quantum algorithm for dissipative nonlinear differential equations

    Jin-Peng Liu et al. “Efficient quantum algorithm for dissipative nonlinear differential equations.” In: Proceedings of the National Academy of Sciences118.35 (Aug. 2021).issn: 1091-6490.doi:10.1073/ pnas.2026805118

    2021

  5. [5]

    World Scientific, 1991

    Krzysztof Kowalski and Willi-hans Steeb.Nonlinear dynamical systems and Carleman linearization. World Scientific, 1991

    1991

  6. [6]

    Quantum Algorithm for Systems of Linear Equations with Exponentially Improved Dependence on Precision

    Andrew M. Childs, Robin Kothari, and Rolando D. Somma. “Quantum Algorithm for Systems of Linear Equations with Exponentially Improved Dependence on Precision.” In:SIAM Journal on Computing 46.6 (Jan. 2017), pp. 1920–1950.issn: 1095-7111.doi:10.1137/16m1087072

    doi:10.1137/16m1087072 2017

  7. [7]

    A shortcut to an optimal quantum linear system solver

    Alexander M Dalzell. “A shortcut to an optimal quantum linear system solver.” In:arXiv preprint arXiv:2406.12086(2024)

    Pith/arXiv arXiv 2024

  8. [8]

    Quantum linear system solvers: A survey of algorithms and applications

    Mauro ES Morales et al. “Quantum linear system solvers: A survey of algorithms and applications.” In: arXiv preprint arXiv:2411.02522(2024)

    Pith/arXiv arXiv 2024

  9. [9]

    Quantum Algorithm for Linear Systems of Equations

    Aram W. Harrow, Avinatan Hassidim, and Seth Lloyd. “Quantum Algorithm for Linear Systems of Equations.” In:Physical Review Letters103.15 (Oct. 2009).issn: 1079-7114.doi:10.1103/physrevlett. 103.150502

    doi:10.1103/physrevlett 2009

  10. [10]

    Variational quantum linear solver

    Carlos Bravo-Prieto et al. “Variational quantum linear solver.” In:Quantum7 (2023), p. 1188

    2023

  11. [11]

    Randomized Adiabatic Quantum Linear Solver Algorithm with Optimal Com- plexity Scaling and Detailed Running Costs

    David Jennings et al. “Randomized Adiabatic Quantum Linear Solver Algorithm with Optimal Com- plexity Scaling and Detailed Running Costs.” In:PRX Quantum6 (4 Dec. 2025), p. 040373.doi: 10.1103/1xkb-22cc.url:https://link.aps.org/doi/10.1103/1xkb-22cc

    work page doi:10.1103/1xkb-22cc.url:https://link.aps.org/doi/10.1103/1xkb-22cc 2025

  12. [12]

    Quantum Algorithms for Systems of Linear Equations Inspired by Adiabatic Quantum Computing

    Yi ğit Suba ş ı, Rolando D. Somma, and Davide Orsucci. “Quantum Algorithms for Systems of Linear Equations Inspired by Adiabatic Quantum Computing.” In:Phys. Rev. Lett.122 (6 Feb. 2019), p. 060504. doi: 10.1103/PhysRevLett.122.060504.url: https://link.aps.org/doi/10.1103/PhysRevLett. 122.060504

    work page doi:10.1103/physrevlett.122.060504.url: 2019

  13. [13]

    Read the fine print

    Scott Aaronson. “Read the fine print.” In:Nature Physics11.4 (2015), pp. 291–293

    2015

  14. [14]

    EfficientdecompositionoftheCarleman linearized Burgers’ equation

    ReubenDemirdjian,ThomasHogancamp,andDanielGunlycke.“EfficientdecompositionoftheCarleman linearized Burgers’ equation.” In:Phys. Rev. A113 (3 Mar. 2026), p. 032408.doi:10.1103/g27q-r2gk. url:https://link.aps.org/doi/10.1103/g27q-r2gk

    doi:10.1103/g27q-r2gk 2026

  15. [15]

    Variational quantum solutions to the advection–diffusion equation for applications in fluid dynamics

    Reuben Demirdjian et al. “Variational quantum solutions to the advection–diffusion equation for applications in fluid dynamics.” In:Quantum Information Processing21.9 (Sept. 2022).issn: 1573-1332. doi:10.1007/s11128-022-03667-7.url:http://dx.doi.org/10.1007/s11128-022-03667-7

    doi:10.1007/s11128-022-03667-7.url:http://dx.doi.org/10.1007/s11128-022-03667-7 2022

  16. [16]

    Efficient quantum access model for sparse structured matrices using linear combination of “things

    Abeynaya Gnanasekaran and Amit Surana. “Efficient quantum access model for sparse structured matrices using linear combination of “things”.” In:Physical Review A113.2 (2026), p. 022437

    2026

  17. [17]

    Potential quantum advantage for simulation of fluid dynamics

    Xiangyu Li et al. “Potential quantum advantage for simulation of fluid dynamics.” In:Physical Review Research7.1 (2025), p. 013036

    2025

  18. [18]

    An end-to-end quantum algorithm for nonlinear fluid dynamics with bounded quantum advantage

    David Jennings et al. “An end-to-end quantum algorithm for nonlinear fluid dynamics with bounded quantum advantage.” In:arXiv preprint arXiv:2512.03758(2025)

    arXiv 2025

  19. [19]

    Variational Quantum Framework for Nonlinear PDE Constrained Optimization Using Carleman Linearization

    Abeynaya Gnanasekaran, Amit Surana, and Hongyu Zhu. “Variational Quantum Framework for Nonlinear PDE Constrained Optimization Using Carleman Linearization.” In:Quantum Information & Computation25.3 (2025), pp. 260–289.doi:10.2478/qic-2025-0014.url: https://doi.org/10. 2478/qic-2025-0014. 27

    work page doi:10.2478/qic-2025-0014.url: 2025

  20. [21]

    Challenges for quantum computation of nonlinear dynamical systems using linear representations

    Yen Ting Lin et al. “Challenges for quantum computation of nonlinear dynamical systems using linear representations.” In:arXiv preprint arXiv:2202.02188(2022)

    Pith/arXiv arXiv 2022

  21. [22]

    The foundational value of quantum computing for classical fluids

    Sauro Succi et al. “The foundational value of quantum computing for classical fluids.” In:Europhysics Letters(2025)

    2025

  22. [23]

    Quantum Carleman linearization efficiency in nonlinear fluid dynamics

    Javier Gonzalez-Conde et al. “Quantum Carleman linearization efficiency in nonlinear fluid dynamics.” In:Physical Review Research7.2 (2025), p. 023254

    2025

  23. [24]

    Quantum algorithms for nonlinear dynamics: Revisiting Carleman linearization with no dissipative conditions

    Hsuan-Cheng Wu, Jingyao Wang, and Xiantao Li. “Quantum algorithms for nonlinear dynamics: Revisiting Carleman linearization with no dissipative conditions.” In:SIAM Journal on Scientific Computing47.2 (2025), A943–A970

    2025

  24. [25]

    Carleman linearization based efficient quan- tum algorithm for higher order polynomial differential equations

    Amit Surana, Abeynaya Gnanasekaran, and Tuhin Sahai. “Carleman linearization based efficient quan- tum algorithm for higher order polynomial differential equations.” In:arXiv preprint ArXiv:2212.10775 (2022)

    Pith/arXiv arXiv 2022

  25. [26]

    Quantum lattice Boltzmann method for several time steps: A local Carleman linearization algorithm

    Antonio David Bastida Zamora et al. “Quantum lattice Boltzmann method for several time steps: A local Carleman linearization algorithm.” In:Physical Review E113.3 (2026), p. 035307

    2026

  26. [27]

    Quantum carleman linearization of the lattice boltzmann equation with boundary conditions

    Bastien Bakker and Thomas W Watts. “Quantum carleman linearization of the lattice boltzmann equation with boundary conditions.” In:arXiv preprint arXiv:2312.04781(2023)

    Pith/arXiv arXiv 2023

  27. [28]

    A Linear Combination of Unitaries Decomposition for the Laplace Operator

    Thomas Hogancamp, Reuben Demirdjian, and Daniel Gunlycke. “A Linear Combination of Unitaries Decomposition for the Laplace Operator.” In:arXiv preprint arXiv:2601.06370(2026)

    arXiv 2026

  28. [29]

    Hamiltonian simulation using linear combinations of unitary operations

    Andrew M Childs and Nathan Wiebe. “Hamiltonian simulation using linear combinations of unitary operations.” In:arXiv preprint arXiv:1202.5822(2012)

    Pith/arXiv arXiv 2012

  29. [30]

    Tensorized Pauli decomposition algorithm

    Lukas Hantzko, Lennart Binkowski, and Sabhyata Gupta. “Tensorized Pauli decomposition algorithm.” In:Physica Scripta99.8 (2024), p. 085128

    2024

  30. [31]

    Efficient Variational Quantum Linear Solver for Structured Sparse Matrices

    Abeynaya Gnanasekaran and Amit Surana. “Efficient Variational Quantum Linear Solver for Structured Sparse Matrices.” In:2024 IEEE International Conference on Quantum Computing and Engineering (QCE). IEEE, Sept. 2024, pp. 199–210.doi:10.1109/qce60285.2024.00033.url: http://dx.doi. org/10.1109/QCE60285.2024.00033

    doi:10.1109/qce60285.2024.00033.url: 2024

  31. [32]

    Efficient quantum access model for sparse structured matrices using linear combination of things

    Abeynaya Gnanasekaran and Amit Surana. “Efficient quantum access model for sparse structured matrices using linear combination of things.” In:arXiv preprint arXiv:2507.03714(2025)

    Pith/arXiv arXiv 2025

  32. [33]

    Quantum Data Loading for Carleman Linearized Systems: Application to the Lattice-Boltzmann Equation

    Reuben Demirdjian et al. “Quantum Data Loading for Carleman Linearized Systems: Application to the Lattice-Boltzmann Equation.” In:arXiv preprint arXiv:2605.00302(2026)

    Pith/arXiv arXiv 2026

  33. [34]

    Barren plateaus in variational quantum computing

    Martin Larocca et al. “Barren plateaus in variational quantum computing.” In:Nature Reviews Physics (2025), pp. 1–16

    2025

  34. [35]

    Nonlinear dynamics as a ground-state solution on quantum computers

    Albert J Pool et al. “Nonlinear dynamics as a ground-state solution on quantum computers.” In:Physical Review Research6.3 (2024), p. 033257

    2024

  35. [36]

    Hierarchical multigrid ansatz for variational quantum algorithms

    Christo Meriwether Keller et al. “Hierarchical multigrid ansatz for variational quantum algorithms.” In: ISC High Performance 2024 Research Paper Proceedings (39th International Conference). Prometeus GmbH. 2024, pp. 1–11

    2024

  36. [37]

    Springer Nature, 2024

    Sheldon Axler.Linear algebra done right. Springer Nature, 2024

    2024

  37. [38]

    2015.url:https://algassert.com/circuits/ 2015/06/05/Constructing-Large-Controlled-Nots.html

    Craig Gidney.Constructing Large Controlled Nots. 2015.url:https://algassert.com/circuits/ 2015/06/05/Constructing-Large-Controlled-Nots.html. (accessed: 04.13.2026)

    2015

  38. [39]

    2024.url: https://egrettathula.wordpress.com/2024/07/ 28/quantum-incrementer/

    Egretta Thula.Quantum Incrementer. 2024.url: https://egrettathula.wordpress.com/2024/07/ 28/quantum-incrementer/. (accessed: 04.13.2026)

    2024

  39. [40]

    Cambridge University Press, Apr

    John Watrous.The Theory of Quantum Information. Cambridge University Press, Apr. 2018.isbn: 9781107180567.doi:10.1017/9781316848142.url:http://dx.doi.org/10.1017/9781316848142

    work page doi:10.1017/9781316848142.url:http://dx.doi.org/10.1017/9781316848142 2018

  40. [41]

    Tutorial on the quantikz package

    Alastair Kay. “Tutorial on the quantikz package.” In:arXiv preprint arXiv:1809.03842(2018). 28

    arXiv 2018

  41. [42]

    2024.doi:10.48550/arXiv.2405.08810

    Ali Javadi-Abhari et al.Quantum computing with Qiskit. 2024.doi:10.48550/arXiv.2405.08810 . arXiv:2405.08810 [quant-ph]

    Pith/arXiv arXiv doi:10.48550/arxiv.2405.08810 2024

  42. [43]

    Expressibility and Entangling Capability of Parameterized Quantum Circuits for Hybrid Quantum-Classical Algorithms

    Sukin Sim, Peter D. Johnson, and Alán Aspuru-Guzik. “Expressibility and Entangling Capability of Parameterized Quantum Circuits for Hybrid Quantum-Classical Algorithms.” In:Advanced Quantum Technologies2.12 (2019), p. 1900070.doi: https : / / doi . org / 10 . 1002 / qute . 201900070. eprint: https://advanced.onlinelibrary.wiley.com/doi/pdf/10.1002/qute.20...

    doi:10.1002/qute.201900070 2019

  43. [44]

    Hierarchical Learning for Training Large- Scale Variational Quantum Circuits

    Hrant Gharibyan, Vincent Paul Su, and Hayk Tepanyan. “Hierarchical Learning for Training Large- Scale Variational Quantum Circuits.” In:2024 International Conference on Machine Learning and Applications (ICMLA). 2024, pp. 1810–1814.doi:10.1109/ICMLA61862.2024.00279

    doi:10.1109/icmla61862.2024.00279 2024

  44. [45]

    On the existence of a normal approximation to the distribution of the ratio of two independent normal random variables

    Eloísa Díaz-Francés and Francisco J. Rubio. “On the existence of a normal approximation to the distribution of the ratio of two independent normal random variables.” In:Statistical Papers54.2 (May 2013), pp. 309–323.issn: 1613-9798.doi:10.1007/s00362-012-0429-2.url: https://doi.org/10. 1007/s00362-012-0429-2

    work page doi:10.1007/s00362-012-0429-2.url: 2013

  45. [46]

    Imaging obstructed ventilation with NMR using inert fluorinated gases

    Dean O. Kuethe et al. “Imaging obstructed ventilation with NMR using inert fluorinated gases.” In: Journal of Applied Physiology88.6 (2000). PMID: 10846046, pp. 2279–2286.doi:10.1152/jappl. 2000 . 88 . 6 . 2279. eprint: https : / / doi . org / 10 . 1152 / jappl . 2000 . 88 . 6 . 2279.url: https : //doi.org/10.1152/jappl.2000.88.6.2279

    work page doi:10.1152/jappl 2000

  46. [47]

    Academic press, 2014

    Dimitri P Bertsekas.Constrained optimization and Lagrange multiplier methods. Academic press, 2014

    2014

  47. [48]

    Adam: A method for stochastic optimization

    Diederik P Kingma and Jimmy Ba. “Adam: A method for stochastic optimization.” In:arXiv preprint arXiv:1412.6980(2014)

    Pith/arXiv arXiv 2014

  48. [49]

    Pennylane: Automatic differentiation of hybrid quantum-classical computations

    Ville Bergholm et al. “Pennylane: Automatic differentiation of hybrid quantum-classical computations.” In:arXiv preprint arXiv:1811.04968(2018)

    Pith/arXiv arXiv 2018

  49. [50]

    Available at https://quantum.cloud.ibm.com/docs/en/guides/qpu- information#2q-error-layered(03/09/2026)

    IBM.QPU information. Available at https://quantum.cloud.ibm.com/docs/en/guides/qpu- information#2q-error-layered(03/09/2026)

    2026

  50. [51]

    2026.url: https : / / quantum

    IBM.EstimatorV2. 2026.url: https : / / quantum . cloud . ibm . com / docs / en / api / qiskit - ibm - runtime/estimator-v2. (accessed: 04.13.2026)

    2026

  51. [52]

    How parallel circuit execution can be useful for NISQ computing?

    Siyuan Niu and Aida Todri-Sanial. “How parallel circuit execution can be useful for NISQ computing?” In:2022 Design, Automation & Test in Europe Conference & Exhibition (DATE). IEEE. 2022, pp. 1065– 1070

    2022

  52. [53]

    Demonstration of the trapped-ion quantum CCD computer architecture

    Juan M Pino et al. “Demonstration of the trapped-ion quantum CCD computer architecture.” In:Nature 592.7853 (2021), pp. 209–213

    2021

  53. [54]

    Efficient and scalable quantum walk algorithms via the quantum Fourier transform

    Asif Shakeel. “Efficient and scalable quantum walk algorithms via the quantum Fourier transform.” In: Quantum Information Processing19.9 (2020), p. 323

    2020

  54. [55]

    Small-scale properties from exascale computations of turbulence on a32 7683 periodic cube

    P.K. Yeung et al. “Small-scale properties from exascale computations of turbulence on a32 7683 periodic cube.” In:Journal of Fluid Mechanics1019 (2025), R2.doi:10.1017/jfm.2025.10493

    doi:10.1017/jfm.2025.10493 2025

  55. [56]

    Rise of conditionally clean ancillae for efficient quantum circuit constructions

    Tanuj Khattar and Craig Gidney. “Rise of conditionally clean ancillae for efficient quantum circuit constructions.” In:Quantum9 (2025), p. 1752

    2025

  56. [57]

    Quantum algorithms for general nonlinear dynamics based on the Carleman embedding

    David Jennings et al. “Quantum algorithms for general nonlinear dynamics based on the Carleman embedding.” In:arXiv preprint arXiv:2509.07155(2025)

    Pith/arXiv arXiv 2025

  57. [58]

    Preconditionedblockencodingsforquantumlinearsystems

    LeighLapworthandChristophSünderhauf.“Preconditionedblockencodingsforquantumlinearsystems.” In:Quantum Science and Technology10.4 (2025), p. 045064

    2025

  58. [59]

    Quantum preconditioning method for linear systems problems via Schr\

    Shi Jin et al. “Quantum preconditioning method for linear systems problems via Schr\" odingerization.” In:arXiv preprint arXiv:2505.06866(2025)

    Pith/arXiv arXiv 2025

  59. [60]

    Preconditioning for a variational quantum linear solver

    Aruto Hosaka et al. “Preconditioning for a variational quantum linear solver.” In:arXiv preprint arXiv:2312.15657(2023)

    Pith/arXiv arXiv 2023

  60. [61]

    Quantum computing and preconditioners for hydrological linear systems

    John Golden, Daniel O’Malley, and Hari Viswanathan. “Quantum computing and preconditioners for hydrological linear systems.” In:Scientific Reports12.1 (2022), p. 22285. 29

    2022

  61. [62]

    Toward Efficient End-to-End Quantum Elliptic PDE Solvers: a Multilevel Correction Algorithm for Direct Observable Estimation

    Xiantao Li. “Toward Efficient End-to-End Quantum Elliptic PDE Solvers: a Multilevel Correction Algorithm for Direct Observable Estimation.” In:arXiv preprint arXiv:2606.01270(2026). 30

    Pith/arXiv arXiv 2026