Pith. sign in

REVIEW 4 major objections 5 minor 40 references

Quantum simulation of multiscale linear transport equations via Schr\"odingerization and exponential integrators

T0 review · 4 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read This paper proposes two Hamiltonian simulation algorithms for the multiscale linear transport equation with incoming boundary conditions; both achieve query complexity $\tilde{O}(N_v N_x^2 \log N_x)$, independent of the mean free path…

desk verdict Iterative method's complexity likely survives a corrected p-grid accounting; steady-state method's headline claim is off by a log—worth a serious referee, not a desk reject. read the letter →

arxiv 2507.18970 v1 pith:6VZSD5UH submitted 2025-07-25 quant-ph math.QA

Xiaoyang He , Shi Jin This is my paper · ORCID
classification quant-phmath.QA MSC 81P6865M06
keywords quantumsimulationlineartransportequationSchrödingerizationexponentialintegratorsasymptotic-preservingschemeHamiltonianmultiscaleproblemsincomingboundaryconditions
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

The paper claims that the multiscale linear transport equation, the phase-space model of particles moving through a scattering medium, can be simulated on a quantum computer in $\tilde{O}(N_v N_x^2 \log N_x)$ queries, with the cost independent of the mean free path $\varepsilon$. It offers two Hamiltonian simulation algorithms, one iterative and one steady-state, both built from the Schr\"odingerization method, a variable change that turns a non-Hermitian linear evolution into a Hermitian Schr\"odinger-type equation, combined with an exponential-integrator, asymptotic-preserving discretization. Incoming boundary conditions are included, which earlier quantum solvers for this equation omitted. If the bounds are right, the second algorithm also produces all intermediate time states, making it useful for time-dependent boundary and interface problems.

What carries the argument

Schr\"odingerization: for a linear ODE $du/dt = Au + b$, the change of variables $\hat{u} = e^{-p}u$ extends the system in an auxiliary $p$-direction and rewrites the evolution as $\partial_t \hat{u} = -A_1 \partial_p \hat{u} + iA_2 \hat{u}$ with Hermitian $A_1, A_2$; after a Fourier spectral discretization in $p$ this becomes unitary dynamics $-iH\hat{u}$ with $H = D_\mu \otimes A_1 - I \otimes A_2$, simulable by a Hamiltonian-simulation oracle. The paper pairs this with an exponential-integrator diffusive-relaxation discretization of the transport equation that is asymptotically preserving, a preprocessing step that rescales the odd-parity flux $j$ by $1/N_x$ to bring all matrix elements to $O(1)$, and a Laplace-transform bound on $\|e^{-HT}\|_2$ that fixes the evolution time for the steady-state solver. These pieces jointly deliver the $\tilde{O}(N_v N_x^2 \log N_x)$ query count.

What would settle it

Take a representative one-dimensional test problem, for instance the constant-coefficient case with $N_x = 64$ and $N_t = N_x^2$, compute the spectrum of $A_1$, and ask whether the $p$-domain $[-L,R]$ with $L=R=N_x$ used in the numerical section satisfies $L > \lambda^-_{\max}(A_1) N_t$; if the answer is no, or if the simulation error does not decay like $O(1/N_x)$ when $N_p$ is held at $O(\log N_x)$, then the claimed query complexity has not been established.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is a way of combining three existing ingredients, Schr\"odingerization, exponential-integrator diffusive relaxation schemes, and a preprocessing rescaling of the odd-parity variable, so that the multiscale transport equation can be solved by Hamiltonian simulation at query complexity $\tilde{O}(N_v N_x^2 \log N_x)$. This improves on the previous quantum linear-system solver's $O(N_v^2 N_x^2 \log N_x)$ by a factor of $N_v$ and is exponentially better in $N_x$ than classical solvers. The key technical step in the steady-state variant is a Laplace-transform-based estimate of the matrix exponential norm $\|e^{-HT}\|_2$ that gives a tight evolution time $T = O(N_t \log(1/\delta))$ instead of a crude norm bound. Both algorithms are claimed to be the first to combine Schr\"odingerization with an effective asymptotic-preserving scheme for this equation, and their complexity is claimed to be independent of $\varepsilon$.

Load-bearing premise

The argument's load-bearing premise is that the auxiliary grid used in the unitary transformation can stay logarithmically small even when the simulation time is $O(N_x^2)$; the paper's own domain-size condition suggests the grid may need to grow proportionally to that time.

Editorial extensions

If this is right

  • Both proposed algorithms solve the multiscale transport equation with incoming boundary conditions at $\tilde{O}(N_v N_x^2 \log N_x)$ queries, removing one factor of $N_v$ relative to the earlier quantum linear-system approach and an exponential factor in $N_x$ relative to classical methods.
  • The complexity is independent of the mean free path $\varepsilon$, so the same circuit budget covers kinetic regimes and the diffusion limit.
  • The steady-state variant returns all intermediate time states, making it usable for time-dependent boundary and interface problems where an iterative time-marcher would need repeated solves.
  • Being Hamiltonian simulation rather than a linear-system solver, the algorithms are more compatible with near-term hardware.
  • This is the first combination of Schr\"odingerization with an asymptotic-preserving scheme for multiscale transport, opening a template for other stiff multiscale equations.

Reading between the lines

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

  • A direct test of the quoted complexity would check whether the $p$-domain length required by the paper's own condition $L > \lambda^-_{\max}(A_1) T$ stays bounded when $T = O(N_x^2)$; if it does not, the $\tilde{O}(N_v N_x^2 \log N_x)$ rate would need an extra factor of $N_x^2$.
  • The Laplace-transform estimate of the matrix exponential is not specific to transport; it could sharpen evolution-time bounds in other steady-state quantum linear solvers where naive norm bounds overestimate $T$.
  • Because the discretization is asymptotic-preserving, the same algorithms should remain well-conditioned as $\varepsilon \to 0$; one testable prediction is that measured query counts stay flat as $\varepsilon$ drops from $10^{-1}$ to $10^{-8}$.
  • A natural next test is to replace the Fourier $p$-grid by an adaptive or wavelet representation: if the $p$-direction can be compressed, the $\varepsilon$-independent bound would hold with a smaller constant and less dependence on smoothness assumptions.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper proposes two quantum Hamiltonian simulation algorithms for the one-dimensional multiscale linear transport equation with incoming boundary conditions, combining the Schrödingerization method with exponential-integrator asymptotic-preserving (AP) discretizations. The first algorithm treats the discrete time-marching scheme as a linear iterative system and solves it via Schrödingerization; the second reformulates the all-time linear system as a steady-state problem for an ODE and solves it by Hamiltonian simulation. Both algorithms are claimed to achieve query and gate complexity O~(N_v N_x^2 log N_x), improving on the HHL-based solver of He, Jin and Yu [11] and on classical methods. The paper includes a detailed error analysis based on Laplace transforms, several lemmas on matrix norm estimates, and numerical experiments for three test problems.

Significance. If the complexity claims were correct, the iterative method would provide a polynomial improvement in N_v over the existing HHL-based quantum solver, and both methods would avoid HHL-style deep circuits, which is relevant for near-term hardware. The combination of Schrödingerization with AP schemes for multiscale transport is a natural and potentially useful extension of the authors' prior framework, and the Laplace-transform-based bound for the matrix exponential is a technically interesting contribution. The numerical experiments are consistent with the algorithms' correctness as proof-of-concept. However, the complexity analysis as written contains several load-bearing inconsistencies, so the headline complexity claims are not established.

major comments (4)
  1. [Section 3.1.1; proofs of Theorems 3.1 and 4.1] The p-grid size N_eta is not O(log(1/delta)) as used in the proofs. Section 3.1.1 requires the p-domain to satisfy L,R > lambda_max(A1) T, and Remark 3.2 together with the CFL condition gives T = s_iter = N_t = O(N_x^2). Hence the number of grid points is N_p = (L+R)/Delta p = O(N_x^2 log(1/delta)). The proofs instead set N_eta = O(log(1/delta)) and identify ||D_eta||_max = N_eta. This is inconsistent with the definition D_mu = diag(2π(k - N_p/2)/(R+L)) in Section 3.1.1, from which ||D_mu||_max = pi/Delta p = O(log(1/delta)) independent of N_p. The derivations as written are therefore invalid; they should be redone with the correct Fourier multiplier norm and with N_p entering only through the Hilbert-space dimension m(H).
  2. [Section 4.2.3; Theorem 4.1 proof] The query complexity claimed for the steady-state method is not established even after the Fourier-symbol correction. Section 4.2.3 gives T_stea = O(N_t log(1/delta)) = O(N_x^2 log N_x) for delta = O(1/N_x). With the corrected ||H||_max = O(log(1/delta)), one obtains chi = O(N_v log(1/delta) * T_stea) = O(N_v N_x^2 log^2 N_x). The proof's displayed equality chi = O(N_v N_t N_eta log(1/delta)) = O(N_v N_x^2 log N_x) is arithmetically false when N_eta = O(log(1/delta)). Thus Theorem 4.1's O(N_v N_x^2 log N_x) is unsupported; the best bound derivable from the paper's own estimates has an extra logarithmic factor.
  3. [Lemma 3.1 application in both theorem proofs] Both proofs substitute chi into Lemma 3.1 but ignore the factor log(chi/delta) in the query complexity. Lemma 3.1 states Q_query = O(chi log(chi/delta)/log log(chi/delta)). Even under the paper's own chi = O(N_v N_x^2 log N_x) and delta = O(1/N_x), this yields Q_query = O~(N_v N_x^2 log^2 N_x), not O~(N_v N_x^2 log N_x). The claimed Q_query and Q_gate in Theorems 3.1 and 4.1 are therefore off by at least a logarithmic factor under the paper's own assumptions.
  4. [Abstract and Section 6] The abstract and the conclusions assert that both algorithms achieve O(N_v N_x^2 log N_x) and outperform existing quantum and classical algorithms for this problem. In light of the previous comments, this statement is not supported by the proofs. The paper should report the corrected complexity bounds and adjust the comparison with He et al. [11] accordingly.
minor comments (5)
  1. [Section 2.5] After Eq. (2.20), the paper says that subsequent variables rn and jn represent the post-processed variables, but the hat notation is dropped inconsistently; later sections should clearly state that the matrices A1, A2, B1, B2 in Sections 3 and 4 refer to the preprocessed matrices.
  2. [Section 3.1.1] The sentence 'in which δ is the desired accuracy' appears after the definition of L and R but is not connected to those definitions; this is likely a typo and should be rephrased.
  3. [Section 5] The numerical experiments use L = R = N_x, whereas the theoretical requirement in Section 3.1.1 is L,R ~ N_t for large T; the experiments therefore operate in a regime where the p-grid size does not reflect the asymptotic complexity analysis. A note acknowledging this would be helpful.
  4. [Lemma 4.1] Lemma 4.1 is cited from Hu et al. [13] without proof, yet it plays a central role in the error analysis of the steady-state method. A proof or a more precise statement should be included.
  5. [Notation for \tilde{O}] The paper defines \tilde{O} as ignoring log log terms, but the complexity expressions also drop polylogarithmic factors such as log(chi/delta) and log^{2.5}(chi/delta). The notation should be clarified or the standard convention for \tilde{O} (hiding all polylog factors) should be adopted.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the claimed query bounds rely on independent prior Schrödingerization lemmas and on external Hamiltonian-simulation complexity; the flagged p-grid inconsistency is a correctness gap rather than a circular reduction.

full rationale

The derivation chain in this manuscript does not exhibit definitional circularity. The two algorithms combine the authors' own Schrödingerization framework with exponential-integrator/asymptotic-preserving discretizations of the multiscale transport equation, but the transport application itself is external to the cited prior results. The key cited ingredients, such as the smooth-initialization estimate N_eta = O(log(1/delta)) from [25] and the error relation Lemma 4.1 from [13], are general statements about the Schrödingerization method and about linear ODE steady-state convergence; they are parameter-free with respect to the transport equation and do not already contain the claimed O~(N_v N_x^2 log N_x) query bound. Under the stated review rules, these are independent supporting results rather than circular imports. The central complexity analysis also uses Berry et al.'s external Hamiltonian-simulation query-count lemma, which is not derived from the target result. No fitted parameter is renamed as a prediction, no known empirical pattern is merely relabeled, and no uniqueness claim from the authors' prior work is used to forbid alternatives. The most serious concern identified in the manuscript is an internal consistency issue: Section 3.1.1 requires the Schrödingerization p-domain to satisfy L, R > lambda_max(A1) T, which would make N_p grow with T, while the proofs of Theorems 3.1 and 4.1 take N_eta = O(log(1/delta)) independently of T. That would affect the correctness or completeness of the stated complexity bound, but it is a mathematical gap or an unsupported assumption, not a circular step in which a claimed output is equivalent to an input by construction. Therefore no circularity is established.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

No numbers are fitted to data, so free_parameters is empty. The central complexity claim rests on several external results: the AP exponential-integrator discretization [32], the Schrödingerization framework [29,30], the sparse Hamiltonian simulation theorem [3], and the grid-dimension bound N_eta=O(log(1/delta)) from [25]. The last of these is applied with evolution time T=N_t=O(N_x^2), although the paper's own Section 3.1.1 requires the p-domain length to scale with T; this is the weakest unproved premise. No new physical entities are invented.

assumptions (6)
  • domain assumption The exponential-integrator AP diffusive relaxation scheme of [32] (Eqs. (2.4)-(2.9)) is uniformly stable and asymptotically preserves the epsilon to 0 diffusion limit.
    Used in Section 2 to derive the discrete marching system (2.15); if the AP property or CFL stability fails, the time-marching system that both quantum algorithms solve is not a valid discretization.
  • domain assumption The Schrödingerization method [29,30] maps the non-Hermitian ODE du/dt=Au+b to unitary dynamics on an auxiliary p-grid, and recovery of u at p=0 or p>=lambda_max(A1)t is valid.
    Section 3.1 uses this to construct H_iter_total and H_stea_total; the whole Hamiltonian simulation encoding depends on this mapping.
  • standard math The p-domain length must satisfy L>lambda^-_max(A1)T and R>lambda^+_max(A1)T (Section 3.1.1).
    This is an internal requirement of the method; it implies N_p=(L+R)/Delta p scales with the evolution time T. The paper does not carry this scaling into the complexity count.
  • ad hoc to paper N_eta=O(log(1/delta)) for smooth initialization, as stated in Remark 3.1 and cited from [25], can be used with T=N_t=O(N_x^2).
    Theorem 3.1 and 4.1 use this to eliminate the p-grid dimension from the dominant complexity; it is not reconciled with the L,R ~ T requirement in Section 3.1.1, so it is an unproved premise introduced to arrive at the claimed result.
  • standard math Berry et al.'s sparse Hamiltonian simulation bound (Lemma 3.1) with sparsity S=O(N_v), max element N_eta, and evolution time N_t or T_stea=N_t log(1/delta) applies to H_iter_total and H_stea_total.
    This external theorem is the machinery that converts Hamiltonian simulation cost into query complexity; its hypotheses (sparsity, norm, time) are asserted in the proofs of Theorems 3.1 and 4.1.
  • standard math For the steady-state method, -H has eigenvalues -1 so the ODE (4.4) has a unique steady state; Lemma 4.2 and the Laplace-transform representation of e^{-HT} provide the convergence-time bound.
    Section 4.1.1 and Appendix B rely on this to justify T_stea=O(N_t log(1/delta)).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Quantum simulation of multiscale linear transport equations via Schr\"odingerization and exponential integrators." pith.science (2026). https://pith.science/paper/6VZSD5UH

@misc{pith2026250718970,
  author       = {Pith},
  title        = {Pith review of: Quantum simulation of multiscale linear transport equations via Schr\"odingerization and exponential integrators},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6VZSD5UH}},
  note         = {Machine review of arXiv:2507.18970}
}
abstract

In this paper, we present two Hamiltonian simulation algorithms for multiscale linear transport equations, combining the Schr\"odingerization method [S. Jin, N. Liu and Y. Yu, Phys. Rev. Lett, 133 (2024), 230602][S. Jin, N. Liu and Y. Yu, Phys. Rev. A, 108 (2023), 032603] and exponential integrator while incorporating incoming boundary conditions. These two algorithms each have advantages in terms of design easiness and scalability, and the query complexity of both algorithms, $\mathcal{O}(N_vN_x^2\log N_x)$, outperforms existing quantum and classical algorithms for solving this equation. In terms of the theoretical framework, these are the first quantum Hamiltonian simulation algorithms for multiscale linear transport equation to combine the Schr\"odingerization method with an effective asymptotic-preserving schemes, which are efficient for handling multiscale problems with stiff terms.

Figures

Figures reproduced from arXiv: 2507.18970 by the authors.

Figure 1
Figure 1. Solution at time t = 0.05 for Problem I using the iterative method, with parameters ∆x = 1/10, ∆t = ∆x 2 and Np = 27 . 6 Conclusions and Discussions In this paper, we propose two Hamiltonian simulation algorithms for multiscale linear transport equations, both of which utilize Schr¨odingerization and exponential integrator methods and consider the incoming boundary condition. Our algorithmic framework differs from t… view at source ↗
Figure 2
Figure 2. Solution at time t = 0.05 for Problem I using the steady-state solution method, with parameters ∆x = 1/10, ∆t = 10/11∆x 2 and Np = 27 . 25 [PITH_FULL_IMAGE:figures/full_fig_p025_2.png] view at source ↗
Figure 3
Figure 3. Solution at time t = 0.1 for Problem II using the iterative method, with parameters ∆x = 1/10, ∆t = ∆x 2 and Np = 210 . 26 [PITH_FULL_IMAGE:figures/full_fig_p026_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Solution at time t = 0.1 for Problem II using the steady-state solution method, with parameters ∆x = 1/10, ∆t = 10/11∆x 2 and Np = 29 . 27 [PITH_FULL_IMAGE:figures/full_fig_p027_4.png]
Figure 5
Figure 5. Figure 5: Solution at time t = 0.05 for Problem III using the iterative method, with parameters ∆x = 1/10, ∆t = ∆x 2 and Np = 27 . 28 [PITH_FULL_IMAGE:figures/full_fig_p028_5.png]
Figure 6
Figure 6. Figure 6: Solution at time t = 0.05 for Problem III using the steady-state solution method, with parameters ∆x = 1/10, ∆t = 10/11∆x 2 and Np = 27 . 29 [PITH_FULL_IMAGE:figures/full_fig_p029_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

40 extracted references · 31 canonical work pages

  1. [11]

    Time complexity analysis of quantum difference methods for the multiscale transport equations

    Xiaoyang He, Shi Jin, and Yue Yu. Time complexity analysis of quantum difference methods for the multiscale transport equations. East Asian Journal on Applied Mathematics , 13:717–739, 2023

    work page 2023

  2. [1]

    Childs, and Lin Lin

    Dong An, Andrew M. Childs, and Lin Lin. Linear combination of hamiltonian simulation for nonunitary dynamics with optimal state preparation cost. Physical Review Letters, 131:150603, 30 2023

    work page 2023

  3. [2]

    D. W. Berry, A. M. Childs, A. Ostrander, and G. Wang. Quantum algorithm for linear dif- ferential equations with exponentially improved dependence on precision. Comm. Math. Phys. , 356(3):1057–1081, 2017

    work page 2017

  4. [3]

    Berry, Andrew M

    Dominic W. Berry, Andrew M. Childs, and Robin Kothari. Hamiltonian simulation with nearly optimal dependence on all parameters. 2015 IEEE 56th Annual Symposium on Foundations of Computer Science, pages 792–809, 2015

    work page 2015

  5. [4]

    Quantum simulation for time-dependent hamiltonians-with applications to non-autonomous ordinary and partial differential equations

    Yu Cao, Shi Jin, and Nana Liu. Quantum simulation for time-dependent hamiltonians-with applications to non-autonomous ordinary and partial differential equations. Journal of Physics A, 58:155304, 2025

    work page 2025

  6. [5]

    Case and Paul Frederick Zweifel

    Kenneth M. Case and Paul Frederick Zweifel. Linear Transport Theory. Addison-Wesley, Reading, MA, 1967

    work page 1967

  7. [6]

    Radiative Transfer

    Subrahmanyan Chandrasekhar. Radiative Transfer. Dover, New York, 1960

    work page 1960

  8. [7]

    A. M. Childs, R. Kothari, and R. D. Somma. Quantum algorithm for systems of linear equations with exponentially improved dependence on precision. SIAM Journal on Scientific Computing , 46(6):1920–1950, 2017

    work page 1920

Show all 40 references
  1. [8]

    P. C. S. Costa, D. An, Y. A. Sanders, Y. Su, R. Babbush, and D. W. Berry. Optimal scaling quantum linear systems solver via discrete adiabatic theorem. arXiv:2111.08152, 2021

    2021 arXiv

  2. [9]

    Quantum simulation of a class of highly-oscillatory transport equations via schr¨ odingerisation.arXiv:2501.10176, 2025

    Anjiao Gu and Shi Jin. Quantum simulation of a class of highly-oscillatory transport equations via schr¨ odingerisation.arXiv:2501.10176, 2025

    2025 arXiv

  3. [10]

    Quantum algorithm for linear systems of equations

    Aram Wettroth Harrow, Avinatan Hassidim, and Seth Lloyd. Quantum algorithm for linear systems of equations. Physical Review Letters, 103:150502, 2009

    2009

  4. [12]

    Asymptotic-preserving schemes for multiscale hyperbolic and kinetic equations

    Jingwei Hu, Shi Jin, and Qin Li. Asymptotic-preserving schemes for multiscale hyperbolic and kinetic equations. In R´ emi Abgrall and Chi-Wang Shu, editors,Handbook of Numerical Methods for Hyperbolic Problems, volume 18 of Handbook of Numerical Analysis, pages 103–129. Elsevier, 2017

    2017

  5. [13]

    Quantum algorithms for multiscale partial differential equations

    Junpeng Hu, Shi Jin, and Lei Zhang. Quantum algorithms for multiscale partial differential equations. Multiscale Modeling & Simulation , 22(3):1030–1067, 2024

    2024

  6. [14]

    Schr¨ odingerization-based quantum implicit-explicit schemes: From time-dependent pdes to multiscale equations, 2025

    Qitong Hu, Xiaoyang He, Xiao-Dong Zhang, and Shi Jin. Schr¨ odingerization-based quantum implicit-explicit schemes: From time-dependent pdes to multiscale equations, 2025. 31

    2025

  7. [15]

    Fundamental patterns of signal propagation in complex net- works

    Qitong Hu and Xiao-Dong Zhang. Fundamental patterns of signal propagation in complex net- works. Chaos, 34:013149, 2024

    2024

  8. [16]

    Information transfer pathways: Signal propagation in complex global topologies

    Qitong Hu and Xiao-Dong Zhang. Information transfer pathways: Signal propagation in complex global topologies. Physica Scripta, 99(7):075206, 2024

    2024

  9. [17]

    Key motifs searching in complex dynamical systems

    Qitong Hu and Xiao-Dong Zhang. Key motifs searching in complex dynamical systems. Physica D, 469:134318, 2024

    2024

  10. [18]

    S. Jin, N. Liu, and Y. Yu. Time complexity analysis of quantum difference methods for linear high dimensional and multiscale partial differential equations. Journal of Computational Physics , 471:111641, 2022

    2022

  11. [19]

    Quantum simulation for partial differential equations with physical boundary or interface conditions

    Shi Jin, Xiantao Li, Nana Liu, and Yue Yu. Quantum simulation for partial differential equations with physical boundary or interface conditions. Journal of Computational Physics , 498:112707, 2024

    2024

  12. [20]

    Quantum simulation of discrete linear dynamical systems and simple iterative methods in linear algebra

    Shi Jin and Nana Liu. Quantum simulation of discrete linear dynamical systems and simple iterative methods in linear algebra. Proceedings of the Royal Society A , 480:20230370, 2024

    2024

  13. [21]

    Quantum simulation of discrete linear dynamical systems and simple iterative methods in linear algebra

    Shi Jin and Nana Liu. Quantum simulation of discrete linear dynamical systems and simple iterative methods in linear algebra. Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences, 480(2292), 2024

    2024

  14. [22]

    Quantum simulation for quantum dynamics with artificial boundary conditions

    Shi Jin, Nana Liu, Xiantao Li, and Yue Yu. Quantum simulation for quantum dynamics with artificial boundary conditions. SIAM Journal on Scientific Computing , 46:B403–B421, 2024

    2024

  15. [23]

    Quantum simulation of maxwell’s equations via schr¨ odingerisation

    Shi Jin, Nana Liu, and Chuwen Ma. Quantum simulation of maxwell’s equations via schr¨ odingerisation. ESAIM: Mathematical Modelling and Numerical Analysis , 58:1853–1879, 2023

    2023

  16. [24]

    On schr¨ odingerization based quantum algorithms for linear dynamical systems with inhomogeneous terms

    Shi Jin, Nana Liu, and Chuwen Ma. On schr¨ odingerization based quantum algorithms for linear dynamical systems with inhomogeneous terms. arXiv:2402.14696, 2024

    2024 arXiv

  17. [25]

    On the schr¨ odingerization method for linear non- unitary dynamics with optimal dependence on matrix queries

    Shi Jin, Nana Liu, Chuwen Ma, and Yue Yu. On the schr¨ odingerization method for linear non- unitary dynamics with optimal dependence on matrix queries. arXiv:2505.00370, 2025

    2025

  18. [26]

    Quantum preconditioning method for linear systems problems via schr¨ odingerization.arXiv:2505.06866, 2025

    Shi Jin, Nana Liu, Chuwen Ma, and Yue Yu. Quantum preconditioning method for linear systems problems via schr¨ odingerization.arXiv:2505.06866, 2025

    2025 arXiv

  19. [27]

    Quantum algorithms for stochastic differential equations: A Schr¨ odingerisation approach.Journal of Scientific Computing , 104(2):56, jun 2025

    Shi Jin, Nana Liu, and Wei Wei. Quantum algorithms for stochastic differential equations: A Schr¨ odingerisation approach.Journal of Scientific Computing , 104(2):56, jun 2025

    2025

  20. [28]

    Time complexity analysis of quantum difference methods for linear high dimensional and multiscale partial differential equations

    Shi Jin, Nana Liu, and Yue Yu. Time complexity analysis of quantum difference methods for linear high dimensional and multiscale partial differential equations. Journal of Computational Physics, 471:111641, 2022. 32

    2022

  21. [29]

    Quantum simulation of partial differential equations: Applications and detailed analysis

    Shi Jin, Nana Liu, and Yue Yu. Quantum simulation of partial differential equations: Applications and detailed analysis. Physical Review A, 108:032603, 2023

    2023

  22. [30]

    Quantum simulation of partial differential equations via schr¨ odingerization.Physical Review Letters, 133:230602, 2024

    Shi Jin, Nana Liu, and Yue Yu. Quantum simulation of partial differential equations via schr¨ odingerization.Physical Review Letters, 133:230602, 2024

    2024

  23. [31]

    Quantum simulation of the fokker-planck equation via schr¨ odingerisation.arXiv:2404.13585, 2024

    Shi Jin, Nana Liu, and Yue Yu. Quantum simulation of the fokker-planck equation via schr¨ odingerisation.arXiv:2404.13585, 2024

    2024 arXiv

  24. [32]

    Uniformly accurate diffusive relaxation schemes for multiscale transport equations

    Shi Jin, Lorenzo Pareschi, and Giuseppe Toscani. Uniformly accurate diffusive relaxation schemes for multiscale transport equations. SIAM Journal on Numerical Analysis , 38:913–936, 2000

    2000

  25. [33]

    Quantum circuits for the black-scholes equations via schr¨ odingerisation.arXiv:2505.04304, 2025

    Shi Jin, Zihao Tang, Xu Yin, and Lei Zhang. Quantum circuits for the black-scholes equations via schr¨ odingerisation.arXiv:2505.04304, 2025

    2025 arXiv

  26. [34]

    Quantum simulation of elastic wave equations via schr¨ odingerisation

    Shi Jin and Chundan Zhang. Quantum simulation of elastic wave equations via schr¨ odingerisation. arXiv:2505.18711, 2025

    2025 arXiv

  27. [35]

    A new asymptotic preserving scheme based on micro- macro formulation for linear kinetic equations in the diffusion limit

    Mohammed Lemou and Luc Mieussens. A new asymptotic preserving scheme based on micro- macro formulation for linear kinetic equations in the diffusion limit. SIAM Journal on Scientific Computing, 31:334–368, 2008

    2008

  28. [36]

    Markowich, Christian A

    Peter A. Markowich, Christian A. Ringhofer, and Christian Schmeiser. Semiconductor Equations. Springer, Vienna, 1990

    1990

  29. [37]

    An implicit monte carlo method for rarefied gas dynamics: I

    Lorenzo Pareschi and Russel Caflisch. An implicit monte carlo method for rarefied gas dynamics: I. the space homogeneous case. Journal of Computational Physics , pages 90–116, 1999

    1999

  30. [38]

    Quantum Computing in the NISQ era and beyond

    John Preskill. Quantum Computing in the NISQ era and beyond. Quantum, 2:79, 2018

    2018

  31. [39]

    Leonid Ryzhik, George Papanicolaou, and Joseph B. Keller. Transport equations for elastic and other waves in random media. Wave Motion, 24:327–370, 1996

    1996

  32. [40]

    Subasi and R

    Y. Subasi and R. D. Somma. Quantum algorithms for systems of linear equations inspired by adiabatic quantum computing. Phys. Rev. Lett., 122:060504, 2019. A Essential properties and lemmas First, we present and prove the following eigenvalue perturbation theorem. Lemma A.1. Co...

    2019

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.