REVIEW 4 major objections 5 minor 50 references
Quantum-Accelerated Solution of Nonlinear Equations from Variational Principles
T0 review · 4 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash
Pith's one-line read Recasting $f(u)=\nabla U(u)=0$ as a gradient flow lets this paper solve nonlinear equilibria on a quantum computer via Carleman or pivot-switching linearization, with query cost independent of system size.
desk verdict Genuinely new gradient-flow linearization trick for nonlinear equilibrium, but the paper's accuracy claims are contradicted by its own numerics and the missing error analysis leaves the advantage unproven. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing mechanism is the gradient-flow reformulation $du/dt=-f(u)=-\nabla U(u)$, which converts a static nonlinear algebraic problem into a dynamical one whose Lyapunov decrease guarantees arrival at equilibrium. Carleman linearization then turns that flow into an infinite-dimensional linear system by promoting each tensor power $y_p=u^{\otimes p}$ to an independent variable; truncating the hierarchy at order $P$ gives a finite matrix that can be block-encoded and simulated with Hamiltonian-simulation or quantum linear-system algorithms. Pivot Switching Carleman (PSC) is the same construction with the highest-order rows expanded around a pivot state $s$ rather than around zero, which stabilizes the truncated system when the plain Carleman matrix develops positive real-part eigenvalues. The spectral norm of the block-encoded matrix is what fixes the query complexity, and the tensor-product structure is what makes the qubit count logarithmic in $N$.
What would settle it
For a chain of $N$ nonlinear springs with increasing nonlinearity, measure the smallest truncation order $P$ that keeps the truncated Carleman or PSC solution within a fixed tolerance of the exact equilibrium, and the time $T$ needed to get there; if either $P$ or $T$ grows with $N$, the claimed system-size independence collapses. Concretely, plot $T(\epsilon,N)=\min\{t:\|u(t)-u^*\|<\epsilon\}$ for $N=8,16,32,\dots$ and check whether $T$ remains bounded.
Extended reading notes
Core claim
The central claim is that any equilibrium equation expressible as the gradient of a coercive, locally Lipschitz potential can be solved by simulating its own energy descent: starting from any initial state, the flow $du/dt=-\nabla U(u)$ reaches a point with $\nabla U=0$ in the limit $t\to\infty$. The authors embed this nonlinear flow into an infinite-dimensional linear system by Carleman linearization, in which the variables are tensor powers $u^{\otimes p}$, or by the Pivot Switching Carleman (PSC) variant, which expands the highest-order truncation rows around a pivot state to suppress divergence. Truncating at order $P$ and block-encoding the resulting matrix $A$ yields a quantum linear-dynamics problem with query complexity $O(P^{3/2}(\|b\|+k+a)t+\ln(1/\epsilon))$, independent of the number of masses $N$. The spring, chain, and truss experiments show the truncated linearized dynamics quantitatively reproduce the exact nonlinear equilibria, with PSC widening the stable parameter range.
Load-bearing premise
The whole argument rests on the truncated linear equations continuing to track the true energy-decreasing motion for the full time needed to reach equilibrium, and on that time not growing rapidly with system size or nonlinearity.
Editorial extensions
If this is right
- Any equilibrium problem of the form $\nabla U(u)=0$ with coercive, locally Lipschitz $U$ becomes, in principle, a quantum linear-dynamics task, so the pipeline is not tied to elasticity.
- The query complexity $O(P^{3/2}(\|b\|+k+a)t+\ln(1/\epsilon))$ contains no factor of $N$, so the cost is set by the potential's coefficients, the truncation order, and the evolution time rather than by the number of degrees of freedom.
- The truncated moment vector has dimension at most $2N^P$, implying a qubit count of $O(P\log N)$ and an exponential memory saving relative to storing the state classically.
- The choice of linearization is decisive: plain Carleman truncation diverges in strongly nonlinear regimes, while PSC with a well-chosen pivot restores convergence, so pivot selection directly governs which nonlinearity strengths are solvable.
- The same variational-to-linear-flow construction is claimed to extend to other equilibrium problems in fluids, electromagnetics, and chemical kinetics whenever the potential satisfies the two structural conditions.
Reading between the lines
- An implication the paper leaves implicit: if the time $T$ needed to reach equilibrium grows with $N$ or with nonlinearity, the $N$-independence of the per-query cost would not translate into an end-to-end advantage; the paper sets $T=1$ in its examples and never analyzes how $T$ scales.
- A natural extension not pursued in the paper: update the PSC pivot state adaptively during the flow so it tracks the evolving solution, which could stabilize the truncation in regimes where the fixed heuristic pivot (such as $s=0.01$) fails.
- The gradient-flow viewpoint also suggests a preconditioning-style initialization: start the Carleman system from a cheap classical approximation of the equilibrium rather than from zero, shortening the simulation time and reducing truncation error in the same way preconditioning accelerates quantum linear-system solvers.
- One could stress-test the claim by running the truncated system for a large chain or truss in classical simulation and asking whether the minimal truncation order $P$ needed for fixed accuracy stays bounded as $N$ grows; the paper reports only small examples ($N=8$ for the chain).
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper proposes a quantum algorithm for solving nonlinear equilibrium equations of the form f(u)=∇U(u)=0 by first converting the problem into the gradient-flow dynamics du/dt=-f(u), then applying Carleman or Pivot Switching Carleman (PSC) linearization to obtain an infinite-dimensional linear system, truncating it, and simulating the result with a quantum linear-dynamics solver. The method is demonstrated on a single nonlinear spring, a one-dimensional chain of springs, and a two-dimensional truss, with numerical results compared against exact solutions. The central advertised claim is that truncated linearized dynamics accurately reproduce the equilibrium states and that the query complexity, Eq. (23), is O(P^{3/2}(‖b‖+k+a)t + ln(1/ε)), independent of the system size N. The paper also gives a complexity analysis based on the spectral norm of the linearized matrix and a convergence proof for the gradient flow in Appendix A.
Significance. If the advertised claims could be made rigorous, the paper would contribute a simple and potentially useful connection between variational equilibrium problems and quantum linear-dynamics simulation, with asymptotic memory savings in N. The paper is also honest in places: the authors explicitly call for further investigation of the evolution time t (§III.B) and acknowledge that Carleman linearization diverges in some regimes (§III.A). The numerical benchmarks are validated against externally computed exact solutions, not against the algorithm's own output, and no fitted parameters are used to force agreement. However, the central claims—that truncated linearized dynamics 'accurately reproduce' the equilibrium and that the complexity is N-independent—are not backed by any truncation-error or convergence-time bound. The paper's own Fig. 2(d) shows Carleman truncation diverging for b≥1.2, contradicting the general accuracy claim. Because the missing pieces are precisely the ones needed to convert a heuristic numerical scheme into a quantum algorithm with a stated complexity, the current version does not establish the main advertised result.
major comments (4)
- [§III.A, Eq. (10) and Fig. 2(d)] The paper states that truncated linearized dynamics 'accurately reproduce nonlinear equilibrium states,' but Fig. 2(d) shows that the Carleman-truncated system diverges for b≥1.2, while the true gradient flow converges. The authors attribute this to positive eigenvalues of the system matrix in Eq. (10) and note that the divergence persists for larger truncation order P. This directly contradicts the general claim in the abstract and conclusion that the truncated dynamics successfully capture the nonlinear equilibrium. The PSC variant repairs the numerics only for the chosen pivot s=0.01, and no error bound is provided for PSC either.
- [§III.B, Eq. (23)] The query-complexity estimate O(P^{3/2}(||b||+k+a)t + ln(1/ε)) treats t as an input parameter, but t is actually the time required for the gradient flow (3) to approach the equilibrium state. The paper itself notes, just after Eq. (23), that 'further investigation is required to estimate preferred length of time evolution t.' Under the stated assumptions (coercive, locally Lipschitz U), convergence can be arbitrarily slow; for example, for U(u)=u^4 one has u'~-u^3, giving ||u(t)||~t^{-1/2} and hence t=O(ε^{-2}) for error ε. Without a bound on T in terms of the problem data, the claimed N-independent complexity is not established, and the numerical choice T=1 in all examples is not justified.
- [§III.B, references [47,48]] Eq. (23) is justified by citing Hamiltonian-simulation results, specifically Low and Chuang's qubitization (refs [47,48]), which apply to unitary (Hermitian) evolution. The linearized Carleman matrix A in Eq. (22) is not Hermitian or necessarily normal. Simulating non-unitary linear dynamics with QLSA or with dilation-based Hamiltonian simulation incurs additional costs related to the condition number, the non-normality, or the norm of the dilated operator. The paper does not provide the end-to-end query complexity for the actual non-unitary simulation, so Eq. (23) as stated is not a demonstrated complexity bound for the proposed algorithm.
- [§III.A, Eq. (11)] The PSC linearization is the method that actually works in the paper's examples, but it is not derived or analyzed here; it is only cited to the authors' prior preprint [41]. The matrix in Eq. (11) is given without explanation of how the pivot state s enters the truncation, and the paper states that s=0.01 is chosen without a priori analysis. There is no proof that choosing a small positive s yields an accurate or stable approximation for general problems. Since the central numerical success of the paper rests on this imported construction, a self-contained derivation or a clear statement of its conditions of validity is needed.
minor comments (5)
- [§III.B, Eq. (13)] The equation as printed reads du_i/dt = -k[(ui-u_{i-1})+(ui-u_{i-1})] - a[(ui-u_{i-1})^3+(ui-u_{i-1})^3] + b_i, with the index i-1 repeated in both terms. This is presumably a typo for terms involving u_{i+1} and u_{i-1}, as in Appendix B, Eq. (B3). Please correct.
- [§III.A, Eq. (11)] The matrix entries such as '4as6' and '−175as6' are notationally ambiguous; they should be typeset as 4a s^6, -175a s^6, etc. The same applies to other entries in the PSC matrix.
- [Appendix A] The proof claims that finiteness of ∫_0^∞ ||∇U||^2 dt implies lim_{t→∞} ||∇U|| = 0. This is not true for a general integrable function. The desired conclusion can be obtained by a standard LaSalle invariance argument using the boundedness of the trajectory, but the written proof is incomplete.
- [Appendix D] The text defines l2 = x^2 + y^2, but Eq. (27) uses l2 as if it denotes u^2 + v^2 (the squared displacement difference). Please clarify the definition to avoid confusing the reader.
- [Fig. 1 and §II.B] The abbreviation 'HS' is used in Fig. 1 but is not defined at that point in the text; please introduce it explicitly (Hamiltonian simulation) before first use in the caption or the text.
Circularity Check
No significant circularity: the equilibrium benchmark is external, the linearization tools are standard or cited as method provenance, and the admitted convergence-time/truncation gaps are complexity limitations, not input-equivalent predictions.
full rationale
Walk-through of the derivation chain: (1) The target equilibrium equation f(u*)=∇U(u*)=0 is not assumed inside the linearized dynamics; it is reformulated as du/dt=-f(u) (Eq. 3), and Appendix A proves that steady states of this gradient flow are exactly the equilibria. The final readout y1(T) is therefore an approximation to the flow's state, not the input f relabeled. (2) The Carleman hierarchy (Eq. 9) and PSC matrix (Eq. 11) are either standard constructions or imported from the authors' prior work [41]; the paper does not fit their parameters to the target equilibrium values. The numerical results in Figs. 2-4 are benchmarked against exact solutions obtained independently with classical solvers, so the agreement is externally falsifiable rather than enforced by construction. (3) The complexity estimate Eq. (23) is an upper bound on the norm of the truncated linear system matrix (Appendix C) and counts queries for a prescribed evolution time t; it does not secretly assume the equilibrium value. The paper explicitly acknowledges at Sec. III.B that 'further investigation is required to estimate preferred length of time evolution t,' and the PSC pivot s=0.01 is described as chosen 'without any a priori analysis.' Those are admitted gaps in resource scaling and pivot selection, and they belong on the correctness/complexity ledger rather than the circularity ledger. No self-citation chain is invoked to forbid alternatives or to force the main construction; PSC's provenance is cited but also demonstrated by the present paper's own numerical experiments. I therefore find no circular step meeting the required reduction test and score 0.
Assumptions & free parameters
free parameters (3)
- Pivot state s =
0.01
- Truncation order P =
5
- Simulation time T =
1
assumptions (4)
- domain assumption U is coercive (U→∞ as ||u||→∞) and ∇U is locally Lipschitz
- domain assumption Carleman linearization exactly represents the nonlinear ODE as an infinite-dimensional linear system, and finite truncation at order P has small error for the regimes tested
- standard math Block-encoding of A/α with α=||A||_2 gives query complexity O(αt + ln(1/ε))
- ad hoc to paper Ancilla qubits and block-encoding circuit depth are constant with respect to N
Cite this review
Pith. "Pith review of Quantum-Accelerated Solution of Nonlinear Equations from Variational Principles." pith.science (2026). https://pith.science/paper/JO3GORBL
@misc{pith2026250817606,
author = {Pith},
title = {Pith review of: Quantum-Accelerated Solution of Nonlinear Equations from Variational Principles},
year = {2026},
howpublished = {\url{https://pith.science/paper/JO3GORBL}},
note = {Machine review of arXiv:2508.17606}
}
read the original abstract
Nonlinear equilibrium problems derived from variational principles arise throughout physics and engineering, including structural mechanics, fluid dynamics, and electromagnetism. While fault-tolerant quantum algorithms have shown promising advantages for linear systems and linear dynamical simulations, extending quantum acceleration to nonlinear equilibrium problems remains a major challenge. Here we introduce a quantum algorithmic framework for nonlinear equilibrium analysis based on gradient-flow linearization. The key idea is to reformulate equilibrium conditions as nonlinear gradient-flow dynamics and transform the resulting evolution into a linear dynamical system using exact linearization techniques such as Carleman and Pivot Switching Carleman (PSC) linearization. This construction enables nonlinear equilibrium and energy-minimization problems to be addressed using quantum algorithms for linear dynamical simulation. We demonstrate the framework for nonlinear elasticity problems ranging from single nonlinear springs and chain-spring systems to two-dimensional truss structures. The resulting truncated linearized dynamics accurately reproduce nonlinear equilibrium states, while PSC linearization substantially improves stability in regimes where conventional Carleman linearization becomes unreliable. More broadly, our work establishes a connection between variational principles, nonlinear energy minimization, exact linearization, and quantum dynamical simulation. This perspective opens a route toward quantum algorithms for nonlinear physical systems beyond the scope of existing linear-system-based approaches.
Figures
Figures from the paper (1 more)
Reference graph
Works this paper leans on
-
[41]
K. Endo and K. Z. Takahashi, Divergence-free algorithms for solving nonlinear differential equations on quantum computers, arXiv preprint arXiv:2411.16233 (2024)
arXiv 2024
-
[1]
L. D. Landau and E. M. Lifshitz, Course of theoretical physics (Elsevier, 2013)
2013
-
[2]
R. P. Feynman, The principle of least action in quan- tum mechanics, in Feynman’s thesis—a new approach to quantum theory (World Scientific, 2005) pp. 1–69
work page 2005
-
[3]
Hilbert, Die grundlagen der physik
D. Hilbert, Die grundlagen der physik . (erste mit- teilung.), Nachrichten von der Gesellschaft der Wis- senschaften zu G¨ ottingen, Mathematisch-Physikalische Klasse 1915, 395 (1915)
work page 1915
-
[4]
E. Noether, Invariante variationsprobleme, in Gesam- melte Abhandlungen-Collected Papers (Springer, 1983) pp. 231–239. 10
work page 1983
-
[5]
Washizu, Variational Methods in Elasticity and Plas- ticity
K. Washizu, Variational Methods in Elasticity and Plas- ticity. (Pergamon Press, 1968)
work page 1968
-
[6]
P. J. Morrison, Hamiltonian description of the ideal fluid, Reviews of modern physics 70, 467 (1998)
work page 1998
-
[7]
J. D. Jackson, Classical electrodynamics (John Wiley & Sons, 2021)
work page 2021
Show all 50 references
-
[8]
Landau, E
L. Landau, E. Lifshitz, and L. E. Reichl, Statistical physics, part 1 (1981)
1981
-
[9]
Tadmor, A review of numerical methods for nonlinear partial differential equations, Bulletin of the American Mathematical Society 49, 507 (2012)
E. Tadmor, A review of numerical methods for nonlinear partial differential equations, Bulletin of the American Mathematical Society 49, 507 (2012)
2012
-
[10]
D. A. Knoll and D. E. Keyes, Jacobian-free newton– krylov methods: a survey of approaches and applications, Journal of Computational Physics 193, 357 (2004)
2004
-
[11]
De, Uncertainty quantification of locally nonlinear dy- namical systems using neural networks, Journal of Com- puting in Civil Engineering 35, 04021009 (2021)
S. De, Uncertainty quantification of locally nonlinear dy- namical systems using neural networks, Journal of Com- puting in Civil Engineering 35, 04021009 (2021)
2021
-
[12]
P. W. Shor, Algorithms for quantum computation: dis- crete logarithms and factoring, in Proceedings 35th an- nual symposium on foundations of computer science (Ieee, 1994) pp. 124–134
1994
-
[13]
L. K. Grover, A fast quantum mechanical algorithm for database search, in Proceedings of the twenty-eighth an- nual ACM symposium on Theory of computing (1996) pp. 212–219
1996
-
[14]
Linden, A
N. Linden, A. Montanaro, and C. Shao, Quantum vs. classical algorithms for solving the heat equation, Com- munications in Mathematical Physics 395, 601 (2022)
2022
-
[15]
P. C. Costa, S. Jordan, and A. Ostrander, Quantum algo- rithm for simulating the wave equation, Physical Review A 99, 012323 (2019)
2019
-
[16]
A. W. Harrow, A. Hassidim, and S. Lloyd, Quantum al- gorithm for linear systems of equations, Physical review letters 103, 150502 (2009)
2009
-
[17]
Gily´ en, Y
A. Gily´ en, Y. Su, G. H. Low, and N. Wiebe, Quantum singular value transformation and beyond: exponential improvements for quantum matrix arithmetics, in Pro- ceedings of the 51st annual ACM SIGACT symposium on theory of computing (2019) pp. 193–204
2019
-
[18]
P. C. Costa, D. An, Y. R. Sanders, Y. Su, R. Bab- bush, and D. W. Berry, Optimal scaling quantum linear- systems solver via discrete adiabatic theorem, PRX quan- tum 3, 040303 (2022)
2022
-
[19]
Suba¸ sı, R
Y. Suba¸ sı, R. D. Somma, and D. Orsucci, Quantum al- gorithms for systems of linear equations inspired by adi- abatic quantum computing, Physical review letters 122, 060504 (2019)
2019
-
[20]
Montanaro and S
A. Montanaro and S. Pallister, Quantum algorithms and the finite element method, Physical Review A 93, 032324 (2016)
2016
-
[21]
Sinha and P
S. Sinha and P. Russer, Quantum computing algorithm for electromagnetic field simulation, Quantum Informa- tion Processing 9, 385 (2010)
2010
-
[22]
B. D. Clader, B. C. Jacobs, and C. R. Sprouse, Precondi- tioned quantum linear system algorithm, Physical review letters 110, 250504 (2013)
2013
-
[23]
D. W. Berry, A. M. Childs, A. Ostrander, and G. Wang, Quantum algorithm for linear differential equations with exponentially improved dependence on precision, Com- munications in Mathematical Physics 356, 1057 (2017)
2017
-
[24]
D. W. Berry, High-order quantum algorithm for solving linear differential equations, Journal of Physics A: Math- ematical and Theoretical 47, 105301 (2014)
2014
-
[25]
Engel, G
A. Engel, G. Smith, and S. E. Parker, Quantum algo- rithm for the vlasov equation, Physical Review A 100, 062315 (2019)
2019
-
[26]
Novikau, E
I. Novikau, E. Startsev, and I. Y. Dodin, Quantum sig- nal processing for simulating cold plasma waves, Physical Review A 105, 062444 (2022)
2022
-
[27]
Miyamoto, S
K. Miyamoto, S. Yamazaki, F. Uchida, K. Fujisawa, and N. Yoshida, Quantum algorithm for the vlasov simula- tion of the large-scale structure formation with massive neutrinos, Physical Review Research 6, 013200 (2024)
2024
-
[28]
A. Suau, G. Staffelbach, and H. Calandra, Practical quantum computing: Solving the wave equation using a quantum approach, ACM Transactions on Quantum Computing 2, 1 (2021)
2021
-
[29]
Babbush, D
R. Babbush, D. W. Berry, R. Kothari, R. D. Somma, and N. Wiebe, Exponential quantum speedup in simulat- ing coupled classical oscillators, Physical Review X 13, 041041 (2023)
2023
-
[30]
Gonzalez-Conde, ´A
J. Gonzalez-Conde, ´A. Rodr´ ıguez-Rozas, E. Solano, and M. Sanz, Efficient hamiltonian simulation for solving op- tion price dynamics, Physical Review Research 5, 043220 (2023)
2023
-
[31]
S. Jin, N. Liu, and Y. Yu, Quantum simulation of par- tial differential equations via schrodingerisation: techni- cal details, arXiv preprint arXiv:2212.14703 (2022)
2022 arXiv
-
[32]
An, J.-P
D. An, J.-P. Liu, and L. Lin, Linear combination of hamil- tonian simulation for nonunitary dynamics with opti- mal state preparation cost, Physical Review Letters 131, 150603 (2023)
2023
-
[33]
D. An, A. M. Childs, and L. Lin, Quantum algorithm for linear non-unitary dynamics with near-optimal depen- dence on all parameters, arXiv preprint arXiv:2312.03916 (2023)
2023
-
[34]
J.-P. Liu, H. Ø. Kolden, H. K. Krovi, N. F. Loureiro, K. Trivisa, and A. M. Childs, Efficient quantum al- gorithm for dissipative nonlinear differential equations, Proceedings of the National Academy of Sciences 118, e2026805118 (2021)
2021
-
[35]
Engel, G
A. Engel, G. Smith, and S. E. Parker, Linear embedding of nonlinear dynamical systems and prospects for effi- cient quantum algorithms, Physics of Plasmas28, 062305 (2021)
2021
-
[36]
J.-P. Liu, D. An, D. Fang, J. Wang, G. H. Low, and S. Jordan, Efficient quantum algorithm for non- linear reaction–diffusion equations and energy estima- tion, Communications in Mathematical Physics 404, 963 (2023)
2023
-
[37]
Costa, P
P. Costa, P. Schleich, M. E. Morales, and D. W. Berry, Further improving quantum algorithms for nonlinear dif- ferential equations via higher-order methods and rescal- ing, arXiv preprint arXiv:2312.09518 (2023)
2023 arXiv
-
[38]
Joseph, Koopman–von neumann approach to quantum simulation of nonlinear classical dynamics, Physical Re- view Research 2, 043102 (2020)
I. Joseph, Koopman–von neumann approach to quantum simulation of nonlinear classical dynamics, Physical Re- view Research 2, 043102 (2020)
2020
-
[39]
S. Jin, N. Liu, and Y. Yu, Time complexity analysis of quantum algorithms via linear representations for nonlin- ear ordinary and partial differential equations, Journal of Computational Physics 487, 112149 (2023)
2023
-
[40]
Krovi, Improved quantum algorithms for linear and nonlinear differential equations, Quantum 7, 913 (2023)
H. Krovi, Improved quantum algorithms for linear and nonlinear differential equations, Quantum 7, 913 (2023)
2023
-
[42]
Xue, X.-F
C. Xue, X.-F. Xu, Y.-C. Wu, and G.-P. Guo, Quantum al- 11 gorithm for solving a quadratic nonlinear system of equa- tions, Physical Review A 106, 032427 (2022)
2022
-
[43]
N. A. Nghiem and T.-C. Wei, Quantum algorithm for solving nonlinear algebraic equations, arXiv preprint arXiv:2404.03810 (2024)
2024
-
[44]
Sanavio, R
C. Sanavio, R. Scatamacchia, C. de Falco, and S. Succi, Three carleman routes to the quantum simulation of clas- sical fluids, Physics of Fluids 36 (2024)
2024
-
[45]
Itani and S
W. Itani and S. Succi, Analysis of carleman linearization of lattice boltzmann, Fluids 7, 24 (2022)
2022
-
[46]
Itani, K
W. Itani, K. R. Sreenivasan, and S. Succi, Quantum algo- rithm for lattice boltzmann (qalb) simulation of incom- pressible fluids with a nonlinear collision term, Physics of Fluids 36 (2024)
2024
-
[47]
G. H. Low and I. L. Chuang, Optimal hamiltonian sim- ulation by quantum signal processing, Physical review letters 118, 010501 (2017)
2017
-
[48]
G. H. Low and I. L. Chuang, Hamiltonian simulation by qubitization, Quantum 3, 163 (2019)
2019
-
[49]
E. A. Coddington and N. Levinson, Theory of ordinary differential equations (McGraw-Hill New York, 1955). Appendix A: A simple proof for the arrival to the solution Consider the time evolution equations du dt =−f(u). (A1) as shown in Eq. 3. f satisfies f(u) = ∇uU (A2) where (1...
1955
-
[50]
(C5) ∥F3∥ =∥aV (ST−I)⊗3− (I−S)⊗3 ∥ = a∥V (ST−I)⊗3− (I−S)⊗3 ∥≤ 2 √ 10a
(C3) We also evaluate the upper bound of the spectral norm of F0,F 1,F 3 as follows: ∥F0∥ =∥b∥ (C4) ∥F1∥ =∥k(S +ST− 2I)∥ =k∥S +ST− 2I∥≤ 4k. (C5) ∥F3∥ =∥aV (ST−I)⊗3− (I−S)⊗3 ∥ = a∥V (ST−I)⊗3− (I−S)⊗3 ∥≤ 2 √ 10a. (C6) Finally, substituting Eq. (C4-6) into Eq. (C3), we get ∥A∥2 2...
Reviewed August 15, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.