Quantum Domain Decomposition for Preconditioning the Finite Element Method

Elise Fressart; Michel Nowak; Nicole Spillane

arxiv: 2605.26090 · v1 · pith:X72E5MYKnew · submitted 2026-05-25 · 🧮 math.NA · cs.NA· quant-ph

Quantum Domain Decomposition for Preconditioning the Finite Element Method

Elise Fressart , Michel Nowak , Nicole Spillane This is my paper

Pith reviewed 2026-06-29 20:11 UTC · model grok-4.3

classification 🧮 math.NA cs.NAquant-ph

keywords quantum linear systemsdomain decompositionfinite element methodpreconditioningPoisson equationAdditive Schwarzblock encoding

0 comments

The pith

It is feasible to apply quantum domain decomposition to precondition the finite element discretization of the Poisson problem using the two-level Additive Schwarz method.

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

The paper establishes that domain decomposition preconditioners can be realized in the quantum setting for linear systems from finite element discretizations. It provides explicit upper bounds on the block-encoding parameters when the two-level Additive Schwarz preconditioner is applied to the Poisson problem, then uses those bounds to determine the complexity of the resulting quantum linear system solver. A sympathetic reader cares because quantum solvers depend heavily on matrix condition number for their runtime, and domain decomposition offers a classical route to improve that parameter while preserving the structure needed for quantum implementation. The work further supplies details on realizing the required operators.

Core claim

We prove that it is feasible to apply quantum domain decomposition. We provide upper bounds for the block-encoding parameters of the Poisson problem discretized by the finite element method and preconditioned by the two-level Additive Schwarz preconditioner. From these bounds, we deduce the complexity of the quantum linear system solver. We also focus on a particular choice of local solver within the domain decomposition preconditioner and provide details on how the operators are implemented.

What carries the argument

The two-level Additive Schwarz preconditioner, which reduces the condition number of the finite-element matrix by combining local subdomain solves with a coarse-space correction and is then block-encoded for quantum use.

If this is right

Upper bounds exist on the block-encoding parameters for the preconditioned finite-element matrix.
The complexity of the quantum linear system solver follows directly from those bounds.
A concrete choice of local solver inside the preconditioner is analyzed for its quantum costs.
Implementation details are given for constructing the necessary quantum operators.

Where Pith is reading between the lines

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

The same block-encoding strategy could be tested on other elliptic problems discretized by finite elements.
Similar analysis might apply to different families of domain decomposition methods beyond the two-level Additive Schwarz approach.
The operator implementation details could serve as a template for adapting additional classical preconditioners to quantum linear solvers.

Load-bearing premise

The local solvers inside the domain decomposition preconditioner can be realized as quantum operators without making the overall block-encoding parameters scale linearly with global problem size.

What would settle it

An explicit construction or cost analysis showing that any quantum realization of the local subdomain solvers requires resources linear in the global mesh size would invalidate the claimed bounds and complexity.

Figures

Figures reproduced from arXiv: 2605.26090 by Elise Fressart, Michel Nowak, Nicole Spillane.

**Figure 1.** Figure 1: Example of 8 overlapping subdomains (d = 2, N1 = 4, N2 = 2, L = 3, and δ = 2−L). restrictions of certain functions in VL. To make things concrete, we also introduce this interpolation operator in the algebraic formalism. The degrees of freedom (interior vertices) of Ω are numbered from 1 to N . The degrees of freedom (interior vertices) of Ω(i) are a subset of vertices that we can denote by ωi . The restri… view at source ↗

read the original abstract

Even in cases where quantum linear solvers provide significant speedup compared to their classical counterparts, their performance depends on some of the same parameters. In particular, the condition number of the matrix which is to be inverted is a decisive parameter. A well known classical, and now quantum, remedy is to precondition the linear system $A x = b$ by premultiplying it by a matrix $H$ in such a way that the condition number of $HA$ is significantly smaller than the condition number of $A$. In this work, we focus on a family of preconditioners called domain decomposition. First, we prove that it is feasible to apply quantum domain decomposition. We provide upper bounds for the block-encoding parameters of the Poisson problem discretized by the finite element method and preconditioned by the two-level Additive Schwarz preconditioner (one of the most fundamental domain decomposition techniques). From these bounds, we deduce the complexity of the quantum linear system solver. Second, we focus on a particular choice of local solver within the domain decomposition preconditioner by applying recent work by [Deiml and Peterseim, \textit{Math. Comput.}, 2025] on the Bramble--Pasciak--Xu (BPX) preconditioner. Finally, we provide details on how the operators are implemented.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper shows how to block-encode a two-level Additive Schwarz preconditioner with BPX local solvers for quantum linear systems on FEM Poisson problems and derives explicit bounds on the encoding parameters.

read the letter

The core contribution is a concrete translation of the classical two-level Additive Schwarz preconditioner into a quantum block-encoding for the finite-element Poisson problem. They start from the known stability properties of the Additive Schwarz operator, incorporate the recent BPX local solver construction from Deiml and Peterseim, and produce upper bounds on the subnormalization factor and query complexity. From those bounds they recover an overall quantum solver complexity that stays polylog in the mesh size under the stated assumptions.

What the paper does cleanly is lay out the operator-level implementation steps and tie the classical approximation theory directly to the block-encoding costs. That step is not automatic; it requires checking that the local corrections and the coarse-space correction can each be realized without introducing extra factors that grow with the number of subdomains. The abstract indicates they perform this check using the cited BPX analysis.

The remaining question is how tight those bounds remain once the local BPX solvers are turned into actual quantum operators. The stress-test note correctly flags that the translation step could re-introduce dependence on the number of subdomains or the global dimension if the local block-encodings are not constructed carefully. If the full paper supplies explicit circuits or query counts for the local pieces and shows the polylog scaling survives, the claim holds; otherwise the complexity reduction is only conditional. No circularity or fitted parameters appear in the argument as described.

This work is aimed at people already comfortable with both domain-decomposition theory and quantum block-encodings who want a worked example for an elliptic model problem. A reader looking for a practical route to preconditioned quantum PDE solvers will find the explicit bounds and implementation notes useful. It is worth sending to a serious referee because the central construction is new, the bounds are stated in terms of existing classical results, and the question of whether the quantum costs stay controlled is a well-posed one that referees can check against the provided proofs.

Referee Report

1 major / 0 minor

Summary. The manuscript claims to establish the feasibility of quantum domain decomposition preconditioning for the finite-element discretization of the Poisson problem. It derives upper bounds on the block-encoding parameters of the two-level Additive Schwarz preconditioner (with a focus on BPX local solvers from Deiml-Peterseim), uses these bounds to obtain the complexity of the resulting quantum linear-system solver, and supplies implementation details for the operators.

Significance. If the stated upper bounds on block-encoding subnormalization factors and query complexities remain polylogarithmic in the global mesh size once the local BPX solvers are realized as quantum operators, the work would supply a concrete route to condition-number-independent quantum solvers for sparse elliptic problems, extending classical domain-decomposition theory into the quantum setting.

major comments (1)

[Abstract (local solver paragraph)] Abstract (paragraph on local solver choice): the central deduction of quantum-solver complexity rests on the claim that classical BPX stability and approximation properties translate into block-encoding costs that stay sublinear/polylog in global N; no explicit verification or operator-norm argument for this translation is visible, and the stress-test concern therefore lands directly on this step.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the careful reading and for isolating the key technical step that must hold for the claimed complexity. We address the concern directly below.

read point-by-point responses

Referee: Abstract (paragraph on local solver choice): the central deduction of quantum-solver complexity rests on the claim that classical BPX stability and approximation properties translate into block-encoding costs that stay sublinear/polylog in global N; no explicit verification or operator-norm argument for this translation is visible, and the stress-test concern therefore lands directly on this step.

Authors: The translation is carried out explicitly in Sections 3 and 4. Section 3 first establishes block-encoding bounds for the two-level Additive Schwarz operator in the general case, expressing the subnormalization factor in terms of the classical stability and approximation constants of the local solvers. Section 4 then specializes to the BPX local solvers of Deiml-Peterseim. Theorems 4.1 and 4.3 invoke the BPX stability and approximation results (their Theorems 3.1 and 3.2) to bound the operator norms of the local correction operators; the resulting subnormalization factors are shown to be O(polylog N) because the classical constants are themselves independent of the global mesh size. The operator-norm arguments appear in the proofs of those theorems and rely only on the standard finite-element inverse inequalities and the BPX estimates; no additional N-dependent factors are introduced. We are prepared to add a short clarifying paragraph in Section 4 that restates this chain of implications if the referee finds the current cross-references insufficiently explicit. revision: partial

Circularity Check

0 steps flagged

No circularity: bounds derived from external BPX properties and Additive Schwarz analysis

full rationale

The paper's chain starts from the classical two-level Additive Schwarz preconditioner for FEM Poisson, invokes the external Deiml-Peterseim BPX construction (different authors) to select local solvers, then supplies explicit upper bounds on block-encoding parameters before deducing quantum solver complexity. No equation reduces by construction to a fitted input, no self-citation is load-bearing, and the cited BPX result is treated as an independent classical input whose translation to quantum norms is presented as a separate implementation step rather than a tautology. The derivation is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review supplies no concrete free parameters, axioms, or invented entities; all such quantities would be visible only in the full proofs and operator constructions.

pith-pipeline@v0.9.1-grok · 5763 in / 1248 out tokens · 27238 ms · 2026-06-29T20:11:26.579364+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

4 extracted references · 4 canonical work pages

[1]

636–647, https://doi.org/10.4230/LIPIcs.STACS

29th international symposium on theoretical aspects of computer science, Paris, France, February 29th – March 3rd, 2012, Wadern: Schloss Dagstuhl – Leib- niz Zentrum f¨ ur Informatik, 2012, pp. 636–647, https://doi.org/10.4230/LIPIcs.STACS. 2012.636. [2]D. An and L. Lin,Quantum linear system solver based on time-optimal adiabatic quan- tum computing and q...

work page doi:10.4230/lipics.stacs 2012
[2]

Bravo-Prieto, R

[8]C. Bravo-Prieto, R. LaRose, M. Cerezo, Y. Subasi, L. Cincio, and P. J. Coles,Vari- ational Quantum Linear Solver, Quantum, 7 (2023), p. 1188, https://doi.org/10.22331/ q-2023-11-22-1188, https://doi.org/10.22331/q-2023-11-22-1188. [9]L. S. Busaleh, J. Kwon, O. Zang, M. Hassan, and Y. Maday,Mitigating barren plateaus via domain decomposition in variatio...

work page doi:10.22331/q-2023-11-22-1188 2023
[3]

Shao and H

[37]C. Shao and H. Xiang,Quantum circulant preconditioner for a linear system of equations, Phys. Rev. A, 98 (2018), p. 062321, https://doi.org/10.1103/PhysRevA.98.062321, https: //link.aps.org/doi/10.1103/PhysRevA.98.062321. [38]N. Spillane,GenEO spectral coarse spaces in SPD domain decomposition, Numerical Al- gorithms, (2025), https://doi.org/10.1007/s...

work page doi:10.1103/physreva.98.062321 2018
[4]

Xu,Iterative methods by space decomposition and subspace correction, SIAM Rev., 34 (1992), pp

[43]J. Xu,Iterative methods by space decomposition and subspace correction, SIAM Rev., 34 (1992), pp. 581–613, https://doi.org/10.1137/1034116. [44]Y. Yuan, C. Wang, B. Wang, Z.-Y. Chen, M.-H. Dou, Y.-C. Wu, and G.-P. Guo,An improved qft-based quantum comparator and extended modular arithmetic using one ancilla qubit, New Journal of Physics, 25 (2023), p....

work page doi:10.1137/1034116 1992

[1] [1]

636–647, https://doi.org/10.4230/LIPIcs.STACS

29th international symposium on theoretical aspects of computer science, Paris, France, February 29th – March 3rd, 2012, Wadern: Schloss Dagstuhl – Leib- niz Zentrum f¨ ur Informatik, 2012, pp. 636–647, https://doi.org/10.4230/LIPIcs.STACS. 2012.636. [2]D. An and L. Lin,Quantum linear system solver based on time-optimal adiabatic quan- tum computing and q...

work page doi:10.4230/lipics.stacs 2012

[2] [2]

Bravo-Prieto, R

[8]C. Bravo-Prieto, R. LaRose, M. Cerezo, Y. Subasi, L. Cincio, and P. J. Coles,Vari- ational Quantum Linear Solver, Quantum, 7 (2023), p. 1188, https://doi.org/10.22331/ q-2023-11-22-1188, https://doi.org/10.22331/q-2023-11-22-1188. [9]L. S. Busaleh, J. Kwon, O. Zang, M. Hassan, and Y. Maday,Mitigating barren plateaus via domain decomposition in variatio...

work page doi:10.22331/q-2023-11-22-1188 2023

[3] [3]

Shao and H

[37]C. Shao and H. Xiang,Quantum circulant preconditioner for a linear system of equations, Phys. Rev. A, 98 (2018), p. 062321, https://doi.org/10.1103/PhysRevA.98.062321, https: //link.aps.org/doi/10.1103/PhysRevA.98.062321. [38]N. Spillane,GenEO spectral coarse spaces in SPD domain decomposition, Numerical Al- gorithms, (2025), https://doi.org/10.1007/s...

work page doi:10.1103/physreva.98.062321 2018

[4] [4]

Xu,Iterative methods by space decomposition and subspace correction, SIAM Rev., 34 (1992), pp

[43]J. Xu,Iterative methods by space decomposition and subspace correction, SIAM Rev., 34 (1992), pp. 581–613, https://doi.org/10.1137/1034116. [44]Y. Yuan, C. Wang, B. Wang, Z.-Y. Chen, M.-H. Dou, Y.-C. Wu, and G.-P. Guo,An improved qft-based quantum comparator and extended modular arithmetic using one ancilla qubit, New Journal of Physics, 25 (2023), p....

work page doi:10.1137/1034116 1992