betaFeedback

WorksComputer Science

Quantum Block Encodings for Periodic Two-Phase Finite Element Operators: 2D Poisson and 2D Elasticity

Abstract

Quantum algorithms require embedding linear operators into unitaries using block encodings. The costs associated with these block encodings determine whether a quantum speedup is achieved. Efficient shift decomposition encodings have been proposed for 2D homogeneous scalar operators. Here, we present exact block encodings for 2D homogeneous elasticity, 2D two-phase bilinear Poisson, and 2D two-phase elasticity periodic finite-element operators. For 2D homogeneous elasticity, the resulting linear combination of unitaries (LCU) has L=17L = 17 terms for every Poisson ratio νν and every mesh resolution, and the subnormalization is the closed form α=E(33+ν)/[6(1−ν2)]α= E(33+ν)/\bigl[6(1-ν^{2})\bigr], exceeding ∥K∥2\|\mathbf{K}\|_{2} by the resolution-independent factor (33+ν)/24(33+ν)/24. To address two-phase periodic microstructures, we introduce two oracles with distinct roles: a node-to-element oracle, built from cyclic shifts, that carries a nodal index to each of the four incident element indices, and a microstructure-specific material oracle that marks phase membership with a sign. This yields 25 and 57 LCU terms for two-phase Poisson and elasticity, respectively; the latter reduces to 49 at ν=1/3ν= 1/3. The term counts are independent of mesh resolution, volume fraction, and phase contrast. The subnormalizations, provided in closed form, are independent of mesh resolution and volume fraction. In the scalar case, the tight bound α=∥K∥∞α= \|\mathbf{K}\|_\infty is achieved whenever a node lies entirely in the stiffer phase. In the elasticity case, that bound is not attained, since the shear coupling contributes entries of both signs to a row. An open-source implementation is available at https://github.com/UW-ERSL/PyBlockEncode

Introduction

Quantum algorithms for linear systems, differential equations, and matrix functions promise substantial speedups for large-scale computational problems [1, 2, 3, 4, 5]. Modern algorithmic formulations require the underlying operator A\mathbf{A} to be embedded in a larger unitary U\mathcal{U} as a block encoding: an (a+n)(a+n)-qubit unitary whose top-left 2n×2n2^n \times2^n block is A/α\mathbf{A}/\alpha for some subnormalization α≥∥A∥2\alpha\ge\lVert\mathbf{A}\rVert_2. Constructing that encoding is the first step of every modern pipeline. The block encoding cost determines whether a quantum speedup is achieved. A general-purpose encoder working from the entries of A\mathbf{A} alone, such as FABLE [6], costs O(4n)O(4^n) gates at α=2nmax⁡ij∣Aij∣\alpha=2^n\max_{ij}|A_{ij}| for a 2n×2n2^n\times2^n matrix, and the exponential overhead is unavoidable without structure [7].

Operators in computational mechanics over regular meshes typically have structure. A finite element discretization on a regular grid is sparse, assembled from one small element matrix, and expressible in terms of cyclic shifts. A substantial literature, reviewed in § 2, uses this structure for scalar 1D and 2D homogeneous operators.

In this work, we construct exact shift-decomposition-based block encodings of three operators: 2D homogeneous elasticity, 2D two-phase bilinear Poisson, and 2D two-phase elasticity.

§ 2 provides a brief technical background and reviews the literature. § 3 recalls the shift-decomposition encoding of the periodic homogeneous Poisson cell, and § 4 derives the same for elasticity. § 5 introduces the two-phase microstructure, the node-to-element mapping and the material oracle. § 6 and § 7 generalize the two encodings to two phases. § 8 describes the circuit and summarizes the resources, and § 9 documents the open-source implementation and the reproduction of every table reported here. Appendix § 15 draws the four circuits at m=2m=2.

ORCID(s): 0000-0002-9688-9697 (K. Suresh)

Background

Block Encoding and LCU

A unitary U\mathcal{U} acting on a+na+n qubits is an (α,a,ε)(\alpha,a,\varepsilon)-block encoding [4] of an nn-qubit operator A\mathbf{A} if

∥A−α(⟨0∣⊗a⊗I)U(∣0⟩⊗a⊗I)∥2≤ε.(1)\left\| \mathbf{A}-\alpha\left(\langle0|^{\otimes a}\otimes I\right)\mathcal{U}\left(|0\rangle^{\otimes a}\otimes I\right)\right\|_{2}\leq\varepsilon. \tag*{(1)}

We place the system register on the least-significant qubits, so the encoded block is the contiguous top-left corner U[:2n,:2n]\mathcal{U}[:2^{n},:2^{n}]. Setting ε=0\varepsilon=0 gives an exact encoding.

The canonical construction assembles U\mathcal{U} as a linear combination of unitaries (LCU) [8, 5]. Given

A=∑k=0L−1ckUk(2)\mathbf{A}=\sum_{k=0}^{L-1}c_{k}U_{k} \tag*{(2)}

with unitaries UkU_k and real coefficients ckc_k, the subnormalization is α=∑k∣ck∣\alpha=\sum_k|c_k|. For symmetric A\mathbf{A}, and whenever every UkU_k is a generalized permutation, carrying one nonzero entry per row,

∥A∥2≤∥A∥∞=max⁡i∑j∣Aij∣≤α=∑k∣ck∣.(3)\|\mathbf{A}\|_{2}\leq\|\mathbf{A}\|_{\infty}=\max_i\sum_j|A_{ij}|\leq\alpha=\sum_k|c_k|. \tag*{(3)}

Every unitary used in this paper is of that form: each UkU_k has at most one entry in a given row, so no cancellation between terms can push α\alpha below the largest absolute row sum, and ∥A∥∞\|\mathbf{A}\|_{\infty} is a floor on what the decomposition can attain.

The subnormalization α\alpha controls both the post-selection success probability p=∥Ax∥2/α2p=\|\mathbf{A}x\|^2/\alpha^2 at ∥x∥2=1\|x\|_2=1, and the query complexity of downstream routines. The circuit then has three stages: PREP loads ∣ck∣/α\sqrt{|c_k|/\alpha} into the ancilla register, SELECT applies UkU_k controlled on the ancilla state ∣k⟩|k\rangle, and UNPREP uncomputes the ancilla, with the sign of ckc_k absorbed into UkU_k or carried by a second preparation. On the ∣0⟩|0\rangle-ancilla subspace,

(⟨0∣anc⊗I)U(∣0⟩anc⊗I)=A/α,α=∑k∣ck∣.(4)\left(\langle0|_{\mathrm{anc}}\otimes I\right)\mathcal{U}\left(|0\rangle_{\mathrm{anc}}\otimes I\right)=\mathbf{A}/\alpha,\qquad\alpha=\sum_k|c_k|. \tag*{(4)}

Consequently, three quantities determine the cost of block-encoding:

  1. The subnormalization α=∑k∣ck∣\alpha=\sum_k|c_k|, which affects the query complexity linearly and the post-selection probability quadratically.

  2. The Toffoli count of SELECT, which, for a generic implementation, is O(L)\mathcal{O}(L) [9].

  3. The arbitrary-angle rotation count of PREP, which dominates the fault-tolerant cost because each rotation costs O(log⁡1/εsyn) T\mathcal{O}(\log1/\varepsilon_{\mathrm{syn}})\ T gates at precision εsyn\varepsilon_{\mathrm{syn}}, while everything else is Clifford and Toffoli.

The PREP register must address all LL terms, so its width satisfies nprep≥⌈log⁡2L⌉n_{\mathrm{prep}}\geq\lceil\log_2 L\rceil. A dense PREP on nprepn_{\mathrm{prep}} qubits uses 2nprep−12^{n_{\mathrm{prep}}}-1 rotations, whether or not every index carries amplitude.

Pauli Encodings

The common generic approach is LCU expansion in the Pauli basis [10]. The Pauli term count, verified numerically for m=2m=2 to 66, is

LPauli={342m,1D Poisson,9164m,2D Poisson (FE),984m−1,2D elasticity,(2m+1−1)2,2D Poisson, 2-phase,1924m−11⋅2m+1,2D elasticity, 2-phase.(5)L_{\mathrm{Pauli}}= \begin{cases} \frac{3}{4}2^m, & \text{1D Poisson},\\ \frac{9}{16}4^m, & \text{2D Poisson (FE)},\\ \frac{9}{8}4^m-1, & \text{2D elasticity},\\ (2^{m+1}-1)^2, & \text{2D Poisson, 2-phase},\\ \frac{19}{2}4^m-11\cdot2^m+1, & \text{2D elasticity, 2-phase}. \end{cases} \tag*{(5)}

for mm qubits per spatial direction. The two 2-phase counts hold at every contrast, but they are specific to the microstructure of Figure 1; other microstructures may give different counts. The subnormalization grows quadratically in mm for both homogeneous 2D cases. Considerable effort has gone into computing Pauli coefficients faster, through tensorized recursion [11], tree traversal [12], symmetry-restricted enumeration [13] and the fast Walsh–Hadamard transform [14].

Image not extracted

Figure 1. The two-phase cell at m=5m = 5: a square inclusion of side N/2N/2, centered, at volume fraction vf=1/4v_f = 1/4. The inclusion carries χe=1\chi_e = 1 and is left unshaded; the matrix carries χe=0\chi_e = 0.

Beyond standard Pauli decompositions, several alternative matrix encoding strategies have been proposed. Lu et al. [15] benchmarked multiple encoding approaches: parallel Walsh–Hadamard expansions, standard LCU methods, and an SVD-based singular value dilation that reduces any real square matrix to an exact two-term LCU. The SVD dilation attains α=∥A∥2\alpha= \lVert\mathbf{A} \rVert_2, but its two factors are dense rather than generalized permutations. Other general block-encoding frameworks, such as state preparation pairs or oracle-based linear combinations of unitaries [4], similarly shift the computational complexity to deep state-preparation circuits or unmapped oracle operations unless specific structural assumptions are made. Xu and Hu [16] remove the operator encoding entirely; instead they vectorize K\mathbf{K} and evaluate the variational cost functions as inner products through a swap test; the expense moves to state preparation, which they state as an open problem.

Kuklinski et al. [7] make the premise of this paper precise: efficient block encodings require structure.

Structured Block Encoding

On a regular grid, a finite difference (FD) or finite element (FE) discretization of the Poisson or elasticity operator produces a matrix that is sparse and translation invariant: every node carries the same stencil, built from a small set of repeated entries. Structured block encoding is the collection of techniques that use this structure so that the gate and ancilla counts scale as O(poly⁡(m))O(\operatorname{poly}(m)) rather than O(2m)O(2^m), and the subnormalization stays constant, for an mm-qubit operator.

Several families of structured block encoding have been proposed. These include shift decomposition, where we write the operator in closed form as a short sum of shift operators and use that sum in an LCU [8, 17, 18, 19, 20]. A closely related recursion expands the Dirichlet stencil in O(m)O(m) terms built from the ladder operators σ±=(X±iY)/2\boldsymbol{\sigma}_{\pm} = (\mathbf{X} \pm i\mathbf{Y})/2 [21]; Gnanasekaran and Surana [22] generalize this to a basis of such operators and recover the decomposition by unitary completion, at O(m)O(m) terms for the one-dimensional Poisson operator and a SELECT cost that grows with the term count. The sparse-access approach places the nonzero values of a structured operator by addressing them directly, through a pair of oracles [23, 24, 25]. Finally, the element-decomposition method, which is specific to FE, assembles the global stiffness matrix from small element matrices [26, 27].

This paper focuses on the shift decomposition approach and addresses two limitations in the literature. The first is the scalar restriction. Shift decomposition has been developed for scalar operators, and the two constructions that address linear elasticity take a different approach: Hölscher et al. [27] take an LCU over elements, so the term count equals the element count and the subnormalization grows with the mesh, and Danz et al. [25] build the stiffness and mass matrices of an elastic structure from general fixed-point arithmetic oracles. Neither yields a decomposition whose length is independent of resolution.

The second limitation is the homogeneity restriction. Block encodings of variable-coefficient operators do exist: Deiml and Peterseim [28] admit an arbitrary spatially varying coefficient but leave its oracle abstract and the subnormalization grows with the ellipticity contrast; Pechan et al. [29] construct a heterogeneous three-dimensional Poisson encoding from the sparse local structure of the discretized operator, at a subnormalization bounded by the largest conductivity in the cell; Yano and Sato [30] represent a piecewise coefficient profile by a controlled Fourier series, which is approximate at a material interface. What none of them provides is a linear combination of fixed length in closed form, over a material oracle realized exactly.

Table 1 compares the scope of the proposed work against the existing literature.

1D Homogeneous Periodic Poisson

Shift decomposition is easiest to see on a 1D operator. For the 1D Poisson problem on N=2mN = 2^m periodic nodes, discretized using linear finite elements of unit element length, leads to the stiffness matrix

K1D=[2−10⋯−1−12−1⋯0⋮⋱⋱⋱⋮−10⋯−12],(6)\mathbf{K}_{1D} = \begin{bmatrix} 2 & -1 & 0 & \cdots& -1 \\ -1 & 2 & -1 & \cdots& 0 \\ \vdots& \ddots& \ddots& \ddots& \vdots\\ -1 & 0 & \cdots& -1 & 2 \end{bmatrix}, \tag*{(6)}

where the corner entries carry the periodic wrap. This can be expressed compactly as

K1D=circ⁡(−1, 2, −1),(7)\mathbf{K}_{1D} = \operatorname{circ}(-1,\,2,\,-1), \tag*{(7)}

Prior block encodings and the present construction. S scalar, V vector; FD finite difference, FE finite element; Dim spatial dimension, dd any dimension; U uniform, H heterogeneous, 2 two-phase; Y/N whether the work names a public repository. O(1)\mathcal{O}(1) means constant in NN; the term count of a shift construction grows linearly with dd for finite differences and as 3d3^d for tensor-product finite elements. {sup}† also treats non-periodic boundary conditions; {sup}‡ holds for constant coefficients; {sup}§ the framework is stated for any $d$, but the stiffness analysis is carried out in one dimension; {sup}⋆ the two SVD factors are dense and unstructured, and no efficient circuit for them is given. n/an/a marks a quantity that does not apply, a dash one the cited work does not report. Every α\alpha is quoted at unit element length.

workOpFE/FDDimMatApproachLLα\alphacode
Kharazi [17] [17]SFDddUshift{sup}†`O(1)\mathcal{O}(1)O(1)\mathcal{O}(1)Y
Sturm [18] [18]SFDddUshiftO(1)\mathcal{O}(1)O(1)\mathcal{O}(1)N
Hogancamp [20] [20]SFDddUshift{sup}†`O(1)\mathcal{O}(1)O(1)\mathcal{O}(1)N
Mahmud [31] [31]SFDddUshift + optimalityO(1)\mathcal{O}(1)O(1)\mathcal{O}(1)N
Gnanasekaran [22] [22]SFD1§1^{\S}Uladder + completionO(m)\mathcal{O}(m)—N
Alkadri [19] [19]SFEddU, Helement LCUO(1)‡\mathcal{O}(1)^{\ddagger}O(1)‡\mathcal{O}(1)^{\ddagger}N
Deiml (2025) [28]SFEddHFE + BPX, oraclen/a∝\propto contrastY
Pechan [29] [29]SFD3Hsparse accessn/a12kmax⁡12k_{\max}Y
Yano [30] [30]SFDddHdiagonal, Fouriern/aassumedN
Danz [25] [25]VFE1§1^{\S}U, Harithmetic oraclen/a—N
Arora [26] [26]SFE1HVQLS; elementN+2N+2n/aN
Lu [15] [15]Sany2n/acoh. VQLS; SVD2⋆2^{\star}1⋆1^{\star}N
Hölscher [27] [27]VFE22element + GroverO(nel)\mathcal{O}(n_{\mathrm{el}})nelδn_{\mathrm{el}}\deltaY
This workS, VFE22shift + reflectionsO(1)\mathcal{O}(1)O(1)\mathcal{O}(1)Y

Table 1. Prior block encodings and the present construction.

where circ⁡(a,b,c)\operatorname{circ}(a,b,c) carries aa on the sub-diagonal, bb on the diagonal and cc on the super-diagonal, each off-diagonal band wrapping into the opposite corner.

Let ScS_c be the cyclic shift

Sc=(000⋯1100⋯0010⋯0⋮⋱⋮⋮00⋯10)(8)S_c = \begin{pmatrix} 0 & 0 & 0 & \cdots& 1 \\ 1 & 0 & 0 & \cdots& 0 \\ 0 & 1 & 0 & \cdots& 0 \\ \vdots& \ddots& \vdots& & \vdots\\ 0 & 0 & \cdots& 1 & 0 \end{pmatrix} \tag*{(8)}

so that Sc:∣j⟩↦∣(j+1) mod N⟩S_c : \lvert j\rangle\mapsto\lvert(j+1) \bmod N\rangle. It is a permutation matrix, hence unitary, and its adjoint Sc†S_c^\dagger is the backward shift. Thus Equation (6) can be expressed as

K1D=2I−Sc−Sc†.(9)\mathbf{K}_{1D} = 2I - S_c - S_c^\dagger. \tag*{(9)}

This is an exact LCU, and hence a block encoding: three unitaries, coefficients known in closed form, and a subnormalization α=4\alpha= 4 that does not grow with NN. The cost is one increment modulo NN of depth O(m)\mathcal{O}(m) [32, 23]; the controlled ripple-carry increment used here costs 2m−22m-2 Toffoli, mm CNOT and m−1m-1 clean ancillas that are returned to ∣0⟩\lvert0\rangle and reused by every shift in the circuit [33, 34].

Table 2 compares the cost of encoding K1D\mathbf{K}_{1D} using shift decomposition against an exact Pauli expansion of the same operator at m=12m=12. The gap widens in 2D and for heterogeneous cells. For the 1D cell, the chain ∥K∥2≤∥K∥∞≤α\lVert K\rVert_2 \le\lVert K\rVert_\infty\le\alpha collapses to equality, so the shift decomposition is optimal.

quantityPaulishift
terms LL3,0723
PREP qubits122
PREP slots4,0964
SELECT Toffoli3,07248
α\alpha14.04.000
psuccp_{\mathrm{succ}}0.0821.00

Table 2. Periodic 1D finite element Poisson cell at m=12m=12 qubits, 4,096 degrees of freedom. The shift column is measured; the Pauli column is exact in closed form, L=32mL=\frac{3}{2^m}, α=m+2\alpha=m+2 and psucc=16(m+2)2p_{\mathrm{succ}}=\frac{16}{(m+2)^2}, since ∥K1D∥2=4\lVert\mathbf{K}_{1D}\rVert_2=4 at every even NN. This is the one cell in which the comparison is tabulated; the four 2D cells follow the same pattern more sharply, at the term counts of Equation (5).

2D Homogeneous Periodic Poisson

For a 2D periodic Poisson problem consisting of a N×NN \times N grid of bilinear Q4Q_4 elements, the finite element operator separates into 1D factors [35, 36, 19],

K2D,sc=K1D⊗M1D+M1D⊗K1D.(10)\mathbf{K}_{2D,\mathrm{sc}} = \mathbf{K}_{1D} \otimes\mathbf{M}_{1D} + \mathbf{M}_{1D} \otimes\mathbf{K}_{1D}. \tag*{(10)}

where M1D=16circ⁡(1,4,1)\mathbf{M}_{1D}=\frac{1}{6}\operatorname{circ}(1,4,1) is the consistent mass matrix. Throughout this paper, the left factor of a Kronecker product acts on the xx register, the next on the yy register; ⊗\otimes thus lists registers from least to most significant, the reverse of the matrix Kronecker product. Both factors are three-band circulants, so both admit the shift decomposition of § 3.1,

M1D=46I+16Sc+16Sc†.(11)\mathbf{M}_{1D}=\frac{4}{6}I+\frac{1}{6}S_c+\frac{1}{6}S_c^\dagger. \tag*{(11)}

and Equation (10) expands over the nine unitaries {I,Sc,Sc†}⊗2\{I,S_c,S_c^\dagger\}^{\otimes2}. This approach to the periodic scalar cell is due to Sturm and Schillo [18] and, for tensor-product finite elements, to Qu-FEM [19].

Merging the coefficients that land on the same pair of shifts,

K2D,sc=83I⊗I−13∑(Up,Uq)≠(I,I)Up⊗Uq,(12)\mathbf{K}_{2D,\mathrm{sc}}=\frac{8}{3}I\otimes I-\frac{1}{3}\sum_{(U_p,U_q)\ne(I,I)}U_p\otimes U_q, \tag*{(12)}

nine terms with

α2D,sc=83+8⋅13=163.(13)\alpha_{2D,\mathrm{sc}}=\frac{8}{3}+8\cdot\frac{1}{3}=\frac{16}{3}. \tag*{(13)}

As a nodal stencil, Equation (12) is the bilinear nine-point Laplacian, and α2D,sc=∥K2D,sc∥∞\alpha_{2D,\mathrm{sc}}=\lVert\mathbf{K}_{2D,\mathrm{sc}}\rVert_\infty exactly.

The shift decomposition is independent of mesh resolution, against 9164m\frac{9}{16}4^m terms at α=(m2+2m+8)/3\alpha=(m^2+2m+8)/3 for a generic Pauli expansion of the same operator.

2D Homogeneous Elasticity

In this section, we extend the above results to 2D homogeneous periodic elasticity. Ordering rows and columns by displacement component gives the familiar 2×22\times2 block

K2D,el=[KxxKxyKyxKyy].(14)\mathbf{K}_{2D,\mathrm{el}}= \begin{bmatrix} \mathbf{K}_{xx} & \mathbf{K}_{xy}\\ \mathbf{K}_{yx} & \mathbf{K}_{yy} \end{bmatrix}. \tag*{(14)}

The blocks separate [35, 37] as

Kxx=C(K1D⊗M1D+1−ν2M1D⊗K1D),(15)\mathbf{K}_{xx}=C\left(\mathbf{K}_{1D}\otimes\mathbf{M}_{1D}+\frac{1-\nu}{2}\mathbf{M}_{1D}\otimes\mathbf{K}_{1D}\right), \tag*{(15)}
Kyy=C(1−ν2K1D⊗M1D+M1D⊗K1D),(16)\mathbf{K}_{yy}=C\left(\frac{1-\nu}{2}\mathbf{K}_{1D}\otimes\mathbf{M}_{1D}+\mathbf{M}_{1D}\otimes\mathbf{K}_{1D}\right), \tag*{(16)}
Kxy=KyxT=−C1+ν2(G1D⊗G1D),(17)\mathbf{K}_{xy}=\mathbf{K}_{yx}^{\mathsf{T}}=-C\frac{1+\nu}{2}\left(\mathbf{G}_{1D}\otimes\mathbf{G}_{1D}\right), \tag*{(17)}

where C=E/(1−ν2)C=E/(1-\nu^2) and G1D=12circ⁡(−1,0,1)\mathbf{G}_{1D}=\frac{1}{2}\operatorname{circ}(-1,0,1) is the antisymmetric gradient coupling, whose shift decomposition is

G1D=12Sc†−12Sc.(18)\mathbf{G}_{1D}=\frac{1}{2}S_c^\dagger-\frac{1}{2}S_c. \tag*{(18)}

The G1D⊗G1D\mathbf{G}_{1D} \otimes\mathbf{G}_{1D} coupling is the new cross-derivative term. Since G1D\mathbf{G}_{1D} is antisymmetric, G1D⊗G1D\mathbf{G}_{1D} \otimes\mathbf{G}_{1D} is symmetric, so Kxy=Kyx\mathbf{K}_{xy} = \mathbf{K}_{yx} on a periodic homogeneous grid.

Elasticity carries nd=2n_d = 2 degrees of freedom per node, so K2D,el\mathbf{K}_{2D,\mathrm{el}} is 2N2×2N22N^2 \times2N^2 and the 2log⁡2N2 \log_2 N spatial qubits no longer address it. One further qubit is required, the dof qubit ∣d⟩|d\rangle, with d=0d = 0 selecting the xx displacement and d=1d = 1 the yy displacement. It is now the most significant qubit of the system register, following the xx and yy registers.

Equation (14) is the sum of three matrices, a block-diagonal average, a block-diagonal difference, and a purely off-diagonal part. With Kxy=Kyx\mathbf{K}_{xy} = \mathbf{K}_{yx} each is a single Kronecker product,

K2D,el=12[Kxx+Kyy00Kxx+Kyy]+12[Kxx−Kyy00Kyy−Kxx]+[0KxyKxy0],(19)\mathbf{K}_{2D,\mathrm{el}} = \frac{1}{2} \begin{bmatrix} \mathbf{K}_{xx}+\mathbf{K}_{yy} & 0 \\ 0 & \mathbf{K}_{xx}+\mathbf{K}_{yy} \end{bmatrix} + \frac{1}{2} \begin{bmatrix} \mathbf{K}_{xx}-\mathbf{K}_{yy} & 0 \\ 0 & \mathbf{K}_{yy}-\mathbf{K}_{xx} \end{bmatrix} + \begin{bmatrix} 0 & \mathbf{K}_{xy} \\ \mathbf{K}_{xy} & 0 \end{bmatrix}, \tag*{(19)}

each term a single Kronecker product of an N2×N2N^2 \times N^2 spatial block with II, ZZ and XX respectively, so that

K2D,el=Kxx+Kyy2⊗I+Kxx−Kyy2⊗Z+Kxy⊗X.(20)\mathbf{K}_{2D,\mathrm{el}} = \frac{\mathbf{K}_{xx}+\mathbf{K}_{yy}}{2} \otimes I + \frac{\mathbf{K}_{xx}-\mathbf{K}_{yy}}{2} \otimes Z + \mathbf{K}_{xy} \otimes X. \tag*{(20)}

This splits K2D,el\mathbf{K}_{2D,\mathrm{el}} into three Pauli components, each an N2×N2N^2 \times N^2 spatial block tensored with one single-qubit matrix. We can now substitute the shift decompositions of K1D\mathbf{K}_{1D}, M1D\mathbf{M}_{1D} and G1D\mathbf{G}_{1D} into (20) to arrive at the following proposition.

Proposition 1. For the periodic plane-stress Q4Q_4 elasticity operator with Young’s modulus EE and Poisson ratio ν\nu, the global stiffness matrix K2D,el\mathbf{K}_{2D,\mathrm{el}} admits the exact LCU

K2D,el=∑(p,q,r)∈TcpqrUp(x)⊗Uq(y)⊗σr(d),(21)\mathbf{K}_{2D,\mathrm{el}} = \sum_{(p,q,r)\in\mathcal{T}} c_{pqr} U_p^{(x)} \otimes U_q^{(y)} \otimes\sigma_r^{(d)}, \tag*{(21)}

with Up,Uq∈{I,Sc,Sc†}U_p, U_q \in\{I, S_c, S_c^\dagger\}, σr∈{I,Z,X}\sigma_r \in\{I, Z, X\}, and T\mathcal{T} the set of triples carrying a nonzero coefficient, of size ∣T∣=17|\mathcal{T}| = 17 for every ν∈(−1,12)\nu\in(-1,\frac{1}{2}) and every N=2mN = 2^m with m≥2m \ge2. The subnormalization is

α2D,el(ν)=E(33+ν)6(1−ν2),(22)\alpha_{2D,\mathrm{el}}(\nu) = \frac{E(33+\nu)}{6(1-\nu^2)}, \tag*{(22)}

constant in NN.

Proof. Substitute the shift decompositions of K1D\mathbf{K}_{1D}, M1D\mathbf{M}_{1D} and G1D\mathbf{G}_{1D} into the three Pauli components of Equation (20) and merge coefficients landing on the same shift pair. The II and ZZ components expand over the nine pairs {I,Sc,Sc†}⊗2\{I,S_c,S_c^\dagger\}^{\otimes2}, the XX component over the four pairs carrying a shift in both directions, since G1D\mathbf{G}_{1D} has no identity term. Of the nine ZZ entries only the four axis pairs survive the antisymmetric combination K1D⊗M1D−M1D⊗K1D\mathbf{K}_{1D}\otimes\mathbf{M}_{1D}-\mathbf{M}_{1D}\otimes\mathbf{K}_{1D}, giving $9+4+4=17$. The nine pairs are distinct as operators for m≥2m \ge2, where Sc≠Sc†S_c \ne S_c^\dagger, so no further merging occurs, and no coefficient vanishes: the II prefactor carries (3−ν)(3-\nu) and the ZZ and XX prefactors carry (1+ν)(1+\nu), vanishing only at ν=3\nu=3 and ν=−1\nu=-1 respectively, neither of which the plane-stress operator admits. For α\alpha, the merged coefficients of the spatial factors sum in magnitude to 16/316/3, 44 and 11 for the II, ZZ and XX components respectively, and weighting them by the component prefactors C(3−ν)/4C(3-\nu)/4, C(1+ν)/4C(1+\nu)/4 and $C(1+\nu)/2 gives Equation (22). □

Table 3 groups the 17 terms. There are at most four distinct magnitudes. The eight non-identity shift pairs are the eight stencil neighbors: four axis pairs such as (S,I)(S,I) and (I,Sc†)(I,S_c^\dagger), and four diagonal pairs such as (S,S)(S,S) and (S†,S)(S^\dagger,S). The spectral norm is ∥K2D,el∥2=4C\lVert\mathbf{K}_{2D,\mathrm{el}}\rVert_2 = 4C for every ν\nu and every NN, so the ratio α2D,el/∥K2D,el∥2\alpha_{2D,\mathrm{el}}/\lVert\mathbf{K}_{2D,\mathrm{el}}\rVert_2 is a closed form,

α2D,el∥K2D,el∥2=33+ν24,(23)\frac{\alpha_{2D,\mathrm{el}}}{\lVert\mathbf{K}_{2D,\mathrm{el}}\rVert_2} = \frac{33+\nu}{24}, \tag*{(23)}

which is 1.3875 at ν=0.3\nu= 0.3.

The term count L=17L = 17 and the closed form α2D,el(ν)\alpha_{2D,\mathrm{el}}(\nu) hold at every m≥2m \ge2, against 984m−1\frac{9}{8}4^m-1 terms for an exact Pauli expansion of the same operator ((5)).

σr\sigma_rstencil offsets (Up(x),Uq(y))(U_p^{(x)}, U_q^{(y)})#coefficientν=0.3\nu= 0.3
II(I,I)(I,I), the centre12(3−ν)3(1−ν2)\frac{2(3-\nu)}{3(1-\nu^2)}+1.9780
IIall eight neighbours8−3−ν12(1−ν2)-\frac{3-\nu}{12(1-\nu^2)}−0.2473
ZZ(I,S)(I,S), (I,S†)(I,S^\dagger)2+14(1−ν)+\frac{1}{4(1-\nu)}+0.3571
ZZ(S,I)(S,I), (S†,I)(S^\dagger,I)2−14(1−ν)-\frac{1}{4(1-\nu)}−0.3571
XX(S,S†)(S,S^\dagger), (S†,S)(S^\dagger,S)2+18(1−ν)+\frac{1}{8(1-\nu)}+0.1786
XX(S,S)(S,S), (S†,S†)(S^\dagger,S^\dagger)2−18(1−ν)-\frac{1}{8(1-\nu)}−0.1786
α2D,el=∑k∣ck∣\alpha_{2D,\mathrm{el}}=\sum_k |c_k|17E(33+ν)6(1−ν2)\frac{E(33+\nu)}{6(1-\nu^2)}6.0989

Table 3. The 17 LCU terms of the periodic plane-stress Q4Q_4 elasticity operator at E=1E=1, grouped by coefficient. S≡ScS \equiv S_c, S†≡Sc†S^\dagger\equiv S_c^\dagger, and the pair (Up(x),Uq(y))(U_p^{(x)},U_q^{(y)}) names the stencil offset. Summing the distinct magnitudes with their multiplicities gives α2D,el(ν)\alpha_{2D,\mathrm{el}}(\nu) of Equation (22).

Two-Phase Microstructure

We now turn our attention to finite element meshes that carry two materials. We refer to the spatial distribution of the two phases as the microstructure. The periodic two-phase cell is the natural target, because it is the unit cell of computational homogenization [38, 39, 40, 41].

With a second phase, the operator is no longer translation invariant. This leads to two challenges. First, the material is defined element by element, whereas the shift operators act on nodal registers, so the encoding must relate the two indices; § 5.1 constructs this node-to-element mapping from cyclic shifts. Second, the element moduli must enter the linear combination as unitaries. A diagonal modulus matrix is not unitary, and expanding it in a generic basis would tie LL to the microstructure; § 5.2 writes it instead as a combination of the identity and a material reflection, with coefficients determined by E1E_1 and E2E_2. § 6 and § 7 then assemble the two-phase operators from these pieces.

Figure 1 shows the primary example used throughout: a square inclusion of side N/2N/2 at the center of the cell. Each phase carries its own material property. For the Poisson problem, E1E_1 and E2E_2 are the conductivities of the two phases; for elasticity, they are the two Young’s moduli, with a common Poisson ratio ν\nu. Each element ee carries a modulus implied from these two phases,

Ee=E2+(E1−E2)χe,χe∈{0,1},(24)E_e = E_2 + (E_1-E_2)\chi_e,\qquad\chi_e \in\{0,1\}, \tag*{(24)}

where χe\chi_e is the microstructure-specific binary field discussed in § 5.2.

The assumed periodicity of the microstructure links opposite boundaries of the domain. Under this assumption, an N×NN \times N mesh carries N2N^2 elements and N2N^2 distinct nodes. We index both nodes and elements using 2m2m qubits, where N=2mN=2^m. We write ∣e⟩|e\rangle, with e=(ex,ey)e=(e_x,e_y), when the register is read as an element index, and ∣i⟩|i\rangle, with i=(ix,iy)i=(i_x,i_y), when read as a nodal index.

Node-to-Element Mapping

The material field is defined element by element, whereas the shift operators of § 3 act on nodal registers. We therefore define a mapping from one to the other.

Each grid node is touched by four adjacent elements, located at the relative spatial offsets

(a,b)∈E:={(0,0),(−1,0),(0,−1),(−1,−1)}.(25)(a,b) \in\mathcal{E} := \{(0,0),(-1,0),(0,-1),(-1,-1)\}. \tag*{(25)}

Figure 2 illustrates this for the node i=(8,3)i=(8,3).

Image not extracted

Figure 2. The four elements incident on node i=(8,3)i = (8, 3), on the 2×22 \times2 patch surrounding it. Each element carries the index of its own lower-left node, marked by a square, so the element at offset (a,b)(a, b) has index i+(a,b)i + (a, b) and the four indices are (8,3)(8, 3), (7,3)(7, 3), (8,2)(8, 2) and (7,2)(7, 2) as (a,b)(a, b) runs over Equation (25).

The node-to-element mapping S(a,b)S_{(a,b)} carries the nodal register ∣i⟩|i\rangle to the element index e=(i+(a,b)) mod Ne=(i+(a,b))\bmod N, componentwise:

S(a,b)=Sca⊗Scb,(26)S_{(a,b)} = S_c^a \otimes S_c^b, \tag*{(26)}

where ScS_c is the 1D cyclic forward shift of Equation (8) and Sc−1=Sc†S_c^{-1}=S_c^\dagger its backward counterpart. Thus, for example,

S(1,0)=Sc1⊗Sc0=Sc⊗I.(27)S_{(1,0)}=S_c^1\otimes S_c^0=S_c\otimes I. \tag*{(27)}

Since E={0,−1}2\mathcal{E}=\{0,-1\}^2, the mapping factorizes over the two spatial registers, and the four offsets of Equation (25) are generated by one conditional shift per register. The offsets are taken modulo NN, so a node on the last row or column reaches elements in the first, and E\mathcal{E} is the same four offsets at every node with no exception at the mesh ends. Periodicity is also what makes S(a,b)S_{(a,b)} a permutation, and hence unitary. S(a,b)S_{(a,b)} acts purely as a spatial coordinate translation, without referencing the material property or the microstructure geometry.

The Material Oracle and Reflection

The microstructure itself is defined as a binary field on the elements,

χ:{0,…,N−1}2→{0,1},(28)\chi:\{0,\ldots,N-1\}^2\to\{0,1\}, \tag*{(28)}

where

vf=1N2∑eχe,(29)v_f=\frac{1}{N^2}\sum_e\chi_e, \tag*{(29)}

is the volume fraction of phase 1. We write r=E1/E2r=E_1/E_2 for the phase contrast ratio.

Two-phase encoding requires a quantum circuit that evaluates the membership field χ\chi. For the square inclusion in Figure 1 the membership test factorizes into independent conditions on exe_x and eye_y. An element lies inside the inclusion when ex∈[N/4,3N/4)e_x\in[N/4,3N/4) and ey∈[N/4,3N/4)e_y\in[N/4,3N/4). The two leading bits of exe_x, with ex,m−1e_{x,m-1} the most significant, place it in one of four quarters, and exe_x lies in range precisely when ex,m−1⊕ex,m−2=1e_{x,m-1}\oplus e_{x,m-2}=1. Applying the identical condition to eye_y yields the Boolean predicate

χe=(ex,m−1⊕ex,m−2)∧(ey,m−1⊕ey,m−2),(30)\chi_e=(e_{x,m-1}\oplus e_{x,m-2})\land(e_{y,m-1}\oplus e_{y,m-2}), \tag*{(30)}

for every m≥2m\ge2. Note that Equation (30) applies only to the microstructure of Figure 1; other microstructures require their own predicates.

Given the Boolean indicator χe\chi_e, the material phase enters the block encoding through the material reflection oracle RχR_\chi:

Rχ=I−2diag⁡(χ),(31)R_\chi=I-2\operatorname{diag}(\chi), \tag*{(31)}

which acts on the element state ∣e⟩\lvert e\rangle by applying a phase flip conditional on phase membership,

Rχ∣e⟩=(−1)χe∣e⟩.(32)R_\chi\lvert e\rangle=(-1)^{\chi_e}\lvert e\rangle. \tag*{(32)}

The operator RχR_\chi is diagonal, Hermitian, and unitary (Rχ2=IR_\chi^2=I). Because it marks states with a phase rather than a flag register, it requires no additional ancilla qubits and enters linear combinations of unitaries directly. Under Equation (30), RχR_\chi acts on exactly four qubits at every mm, the two leading bits of each coordinate: two CNOT gates compute the two parities in place, one controlled ZZ applies the phase, and two CNOT gates restore the register. No ancilla is needed, because the parities are computed in place on the coordinate registers rather than onto shift-control qubits. For this microstructure the cost is four CNOT and one controlled ZZ at every m≥2m\ge2.

Combining the mapping S(a,b)S_{(a,b)} with the base oracle RχR_\chi yields the four element-wise material reflection operators R(a,b)R_{(a,b)} as seen from node ii. Reading right to left, the nodal index is carried to the element index, the oracle is queried there, and the index is carried back:

R(a,b)=S(a,b)†RχS(a,b)=I−2diag⁡(χi+(a,b))i.(33)R_{(a,b)}=S_{(a,b)}^\dagger R_\chi S_{(a,b)}=I-2\operatorname{diag}\left(\chi_{i+(a,b)}\right)_i. \tag*{(33)}

Writing D(a,b):=diag⁡(Ei+(a,b))iD_{(a,b)}:=\operatorname{diag}\left(E_{i+(a,b)}\right)_i for the modulus of the element at offset (a,b)(a,b) as seen from node ii, the two-phase field of Equation (24) gives

D(a,b)=E1+E22I−E1−E22R(a,b).(34)D_{(a,b)}=\frac{E_1+E_2}{2}I-\frac{E_1-E_2}{2}R_{(a,b)}. \tag*{(34)}

Two-Phase Poisson

Section 3.2 decomposed the homogeneous cell into 1D factors. A variable modulus destroys that separation, so we work instead at the level of the element matrix, where the modulus is a scalar multiplier and the shift structure survives. At unit conductivity the bilinear Q4Q_4 element matrix is

ke=16[4−1−2−1−14−1−2−2−14−1−1−2−14],(35)\mathbf{k}^{e}=\frac{1}{6}\begin{bmatrix}4&-1&-2&-1\\-1&4&-1&-2\\-2&-1&4&-1\\-1&-2&-1&4\end{bmatrix}, \tag*{(35)}

its rows and columns running counterclockwise from the lower-left corner. We index them by the corners, writing ke[c,c′]k^{e}[c,c^{\prime}] with c,c′∈{0,1}2c,c^{\prime}\in\{0,1\}^{2}.

For node (ix,iy)(i_x,i_y), Figure 2 shows the four incident elements and their offsets (a,b)∈E(a,b)\in\mathcal{E}. Each of those elements couples that node to all four of its own nodes, so a second index is needed to enumerate them. The element at offset (a,b)(a,b) has lower-left node (ix,iy)+(a,b)(i_x,i_y)+(a,b), so node (ix,iy)(i_x,i_y) is its corner c=−(a,b)c=-(a,b) and its four nodes are (ix,iy)+(a,b)+c′(i_x,i_y)+(a,b)+c^{\prime} as c′c^{\prime} runs over {0,1}2\{0,1\}^{2}. That node couples to node (ix,iy)+δ(i_x,i_y)+\delta, with δ=(a,b)+c′\delta=(a,b)+c^{\prime}, through the entry ke[c,c′]k^{e}[c,c^{\prime}].

An entry in the row at (ix,iy)(i_x,i_y) and the column at (ix,iy)+δ(i_x,i_y)+\delta is the operator Sc−δx⊗Sc′−δyS_{c}^{-\delta_x}\otimes S_{c^{\prime}}^{-\delta_y}, so the element at (a,b)(a,b) contributes the four-term operator

T(a,b):=∑c′∈{0,1}2ke[c,c′]Sc−δx⊗Sc′−δy,(36)\mathbf{T}_{(a,b)}:=\sum_{c^{\prime}\in\{0,1\}^{2}}k^{e}[c,c^{\prime}]S_{c}^{-\delta_x}\otimes S_{c^{\prime}}^{-\delta_y}, \tag*{(36)}

with δ=(a,b)+c′\delta=(a,b)+c^{\prime} running with the sum. Scaling each element by its modulus gives the assembled operator

K2D,scχ=∑(a,b)∈ED(a,b)T(a,b),(37)\mathbf{K}^{\chi}_{2\mathrm{D},\mathrm{sc}}=\sum_{(a,b)\in\mathcal{E}}D_{(a,b)}\mathbf{T}_{(a,b)}, \tag*{(37)}

with D(a,b)D_{(a,b)} defined in Equation (34).

The four T(a,b)\mathbf{T}_{(a,b)} merge into the nine-term homogeneous operator of Equation (12),

∑(a,b)∈ET(a,b)=K2D,sc,(38)\sum_{(a,b)\in\mathcal{E}}\mathbf{T}_{(a,b)}=\mathbf{K}_{2\mathrm{D},\mathrm{sc}}, \tag*{(38)}

and each carries the same four coefficients up to order, hence the same coefficient sum 43\frac{4}{3}, so the four together sum to 163\frac{16}{3}, which is α2D,sc\alpha_{2\mathrm{D},\mathrm{sc}} of Equation (13).

Substituting Equation (34) into Equation (37) splits the operator into a mean part, which merges by Equation (38), and a contrast part, which does not:

K2D,scχ=E1+E22K2D,sc−E1−E22∑(a,b)∈ER(a,b)T(a,b).(39)\mathbf{K}^{\chi}_{2\mathrm{D},\mathrm{sc}}=\frac{E_1+E_2}{2}\mathbf{K}_{2\mathrm{D},\mathrm{sc}}-\frac{E_1-E_2}{2}\sum_{(a,b)\in\mathcal{E}}R_{(a,b)}\mathbf{T}_{(a,b)}. \tag*{(39)}

The microstructure appears only inside the reflections.

Proposition 2. Equation (39) is an exact LCU of the two-phase periodic Q4Q_4 Poisson operator in which every unitary is a cyclic shift pair, optionally preceded by one material reflection. For every E1≠E2E_1\ne E_2, every χ\chi and every N=2mN=2^m with m≥2m\ge2 it has

L=9+16=25,α2D,scχ=163max⁡(E1,E2).(40)L=9+16=25,\qquad\alpha^{\chi}_{2\mathrm{D},\mathrm{sc}}=\frac{16}{3}\max(E_1,E_2). \tag*{(40)}

At E1=E2E_1=E_2 the second family vanishes and LL collapses to the 9 of Equation (12).

Proof. The first family of Equation (39) is K2D,sc\mathbf{K}_{2\mathrm{D},\mathrm{sc}}, nine terms. The second is four reflections carrying four shifts each, and the sixteen combinations occupy sixteen index slots distinct from each other and from the first family, so nothing merges: L=25L=25. For α2D,scχ\alpha^{\chi}_{2\mathrm{D},\mathrm{sc}}, the first family contributes 12(E1+E2)α2D,sc\frac{1}{2}(E_1+E_2)\alpha_{2\mathrm{D},\mathrm{sc}} and the second 12∣E1−E2∣∑(a,b)∥T(a,b)∥\frac{1}{2}|E_1-E_2|\sum_{(a,b)}\lVert\mathbf{T}_{(a,b)}\rVert, where ∥T(a,b)∥\lVert\mathbf{T}_{(a,b)}\rVert denotes the sum of the coefficient magnitudes of T(a,b)\mathbf{T}_{(a,b)}. Each one ∥T(a,b)∥\lVert\mathbf{T}_{(a,b)}\rVert is 43\frac{4}{3}, so the second sum is 163\frac{16}{3}, as is α2D,sc\alpha_{2\mathrm{D},\mathrm{sc}}, giving α2D,scχ=163[(E1+E2)+∣E1−E2∣]/2\alpha^{\chi}_{2\mathrm{D},\mathrm{sc}}=\frac{16}{3}[(E_1+E_2)+|E_1-E_2|]/2, which is 163max⁡(E1,E2)\frac{16}{3}\max(E_1,E_2). The factor E1−E2E_1-E_2 gives the collapse.

Material contrast introduces no extra subnormalization penalty beyond scaling with the stiffer phase’s modulus. Switching the soft and stiff phases gives the same subnormalization.

This subnormalization reaches the smallest value a shift-based LCU can attain, α2D,scχ=∥K2D,scχ∥∞\alpha^\chi_{2D,\mathrm{sc}} = \lVert\mathbf{K}^\chi_{2D,\mathrm{sc}} \rVert_\infty, whenever the mesh contains at least one node completely surrounded by the stiffer material. Because the matrix infinity norm ∥K2D,scχ∥∞\lVert\mathbf{K}^\chi_{2D,\mathrm{sc}} \rVert_\infty is the maximum absolute row sum across all nodes, a single fully stiffened node raises the matrix norm to its upper limit of 163max⁡(E1,E2)\frac{16}{3}\max(E_1,E_2), which equals α2D,scχ\alpha^\chi_{2D,\mathrm{sc}}.

If no such node exists, as in a fine checkerboard pattern where every node touches a mix of soft and stiff elements, the maximum row sum ∥K2D,scχ∥∞\lVert\mathbf{K}^\chi_{2D,\mathrm{sc}} \rVert_\infty is smaller than 163max⁡(E1,E2)\frac{16}{3}\max(E_1,E_2), making the bound less tight. For example, an 8×88 \times8 checkerboard with a 3:13:1 material contrast yields ∥K2D,scχ∥∞=10.667\lVert\mathbf{K}^\chi_{2D,\mathrm{sc}} \rVert_\infty= 10.667 against α2D,scχ=16\alpha^\chi_{2D,\mathrm{sc}} = 16. The encoding remains exact in all cases, but α2D,scχ\alpha^\chi_{2D,\mathrm{sc}} reflects a worst-case material bound.

Neither L=25L = 25 nor the circuit varies with the contrast; only the PREP angles and α2D,scχ\alpha^\chi_{2D,\mathrm{sc}} do. An exact Pauli expansion of the same operator leads to (2m+1−1)2(2^{m+1}-1)^2 terms at every contrast for the microstructure of Figure 1 (Equation (5)).

Two-Phase Elasticity

For two-phase elasticity, the construction of § 6 applies unchanged once k[c,c′]ek^{e}_{[c,c']} becomes a 2×22 \times2 block. Two things differ, and both belong to elasticity rather than to the method: a fourth dof component, iYiY, appears, and the four element contributions at a node no longer share a sign.

The plane-stress Q4Q_4 element matrix is 8×88 \times8, two degrees of freedom per node. Indexing rows and columns by corners as before, k[c,c′]ek^{e}_{[c,c']} is now the 2×22 \times2 block coupling the two displacement components at the corner-cc node to those at the corner-c′c' node. Each block resolves over the real orthogonal basis {I,Z,X,iY}\{I,Z,X,iY\} of the dof register, the first three as in § 4 and the fourth iY=(01−10)iY = \begin{pmatrix}0 & 1 \\ -1 & 0\end{pmatrix}, which is real and orthogonal and carries the antisymmetric part of each block. Writing [⋅]r[\cdot]_r for the component along σr\sigma_r, the element at offset (a,b)(a,b) contributes

T(a,b):=∑c′∈{0,1}2∑r[k[c,c′]e]rSc−δx⊗Sc−δy⊗σr,(41)\mathbf{T}_{(a,b)} := \sum_{c' \in\{0,1\}^2} \sum_r [k^{e}_{[c,c']}]_r S_c^{-\delta_x} \otimes S_c^{-\delta_y} \otimes\sigma_r, \tag*{(41)}

with c=−(a,b)c = -(a,b) and δ=(a,b)+c′\delta= (a,b) + c' as in Equation (36). The assembly and the split are then Equations (37) and (39) verbatim, with D(a,b)D_{(a,b)} acting on the spatial registers alone:

K2D,elχ=E1+E22K2D,el−E1−E22∑(a,b)∈ER(a,b)T(a,b).(42)\mathbf{K}^\chi_{2D,\mathrm{el}} = \frac{E_1+E_2}{2}\mathbf{K}_{2D,\mathrm{el}} - \frac{E_1-E_2}{2}\sum_{(a,b)\in\mathcal{E}} R_{(a,b)}\mathbf{T}_{(a,b)}. \tag*{(42)}

The two properties that carried § 6 now read differently. The merge still holds, ∑(a,b)T(a,b)=K2D,el\sum_{(a,b)} \mathbf{T}_{(a,b)} = \mathbf{K}_{2D,\mathrm{el}}, the 17 terms of Proposition 1. But each T(a,b)\mathbf{T}_{(a,b)} now carries ten terms rather than four, distributed 2,3,3,22,3,3,2 over its four offsets: the offset at the node itself and the diagonal offset carry {I,X}\{I,X\}, while the two axis offsets carry {I,Z,iY}\{I,Z,iY\}. The reflection family therefore holds 40 terms.

The iYiY component is absent from the merged family and present in the reflection family, and the reason is structural. Merging restores the symmetry Kxy=Kyx\mathbf{K}_{xy} = \mathbf{K}_{yx} on which the three-component rewriting of Equation (20) depends; with a variable modulus each element is weighted separately, the shear block loses that symmetry, and iYiY carries the antisymmetric remainder. Its total weight is

∑(a,b)∑δ∣[T(a,b)]iY∣=∣1−3ν∣1−ν2,(43)\sum_{(a,b)} \sum_{\delta} \left|[\mathbf{T}_{(a,b)}]_{iY}\right| = \frac{|1-3\nu|}{1-\nu^2}, \tag*{(43)}

so the eight iYiY terms vanish at ν=1/3\nu= 1/3 and nowhere else.

Proposition 3. (42) is an exact LCU of the two-phase periodic plane-stress Q4Q_4 operator in which every unitary is a cyclic shift pair tensored with a single-qubit σr∈{I,Z,X,iY}\sigma_r \in\{I,Z,X,iY\}, optionally preceded by one material reflection. For every E1≠E2E_1 \ne E_2, every χ\chi, every ν∈(−1,12)\nu\in(-1,\frac{1}{2}) and every N=2mN = 2^m with m≥2m \ge2,

L=17+40=57,L=49 at ν=1/3,(44)L = 17 + 40 = 57,\qquad L = 49\ \text{at }\nu= 1/3, \tag*{(44)}

and the subnormalization is

α2D,elχ=min⁡(E1,E2)A(ν)+∣E1−E2∣B(ν),(45)\alpha^{\chi}_{2\mathrm{D},\mathrm{el}}=\min(E_1,E_2)A(\nu)+\lvert E_1-E_2\rvert B(\nu), \tag*{(45)}

with A(ν)=α2D,elχA(\nu)=\alpha_{2\mathrm{D},\mathrm{el}}^{\chi} of Equation (22) at E=1E=1 and

B(ν)=69+5ν+6∣1−3ν∣12(1−ν2).(46)B(\nu)=\frac{69+5\nu+6\lvert1-3\nu\rvert}{12(1-\nu^2)}. \tag*{(46)}

At E1=E2E_1=E_2 the second family vanishes and LL collapses to 17.

Proof. The counting is that of Proposition 2 with ten terms per element in place of four, and the eight iYiY terms carry total weight ∣1−3ν∣/(1−ν2)\lvert1-3\nu\rvert/(1-\nu^2) by Equation (43), vanishing at ν=1/3\nu=1/3 and nowhere else, which is the drop to 49. For the subnormalization, summing magnitudes in Equation (42) gives 12(E1+E2)A+12∣E1−E2∣∑(a,b)∥T(a,b)∥\frac{1}{2}(E_1+E_2)A+\frac{1}{2}\lvert E_1-E_2\rvert\sum_{(a,b)}\lVert T_{(a,b)}\rVert, and writing 12(E1+E2)=min⁡(E1,E2)+12∣E1−E2∣\frac{1}{2}(E_1+E_2)=\min(E_1,E_2)+\frac{1}{2}\lvert E_1-E_2\rvert collects the two into Equation (45) with B=(A+∑(a,b)∥T(a,b)∥)/2B=(A+\sum_{(a,b)}\lVert T_{(a,b)}\rVert)/2. Evaluating that sum on the plane-stress element gives Equation (46), whose absolute-value term is half the iYiY weight of Equation (43).

Written out, Equation (45) reads

α2D,elχ=16(1−ν2)[12(E1+E2)(33+ν)+∣E1−E2∣(18+2ν+3∣1−3ν∣)].(47)\alpha^{\chi}_{2\mathrm{D},\mathrm{el}}= \frac{1}{6(1-\nu^2)} \left[ \frac{1}{2}(E_1+E_2)(33+\nu) +\lvert E_1-E_2\rvert(18+2\nu+3\lvert1-3\nu\rvert) \right]. \tag*{(47)}

As in the scalar cell, the reflections are present at every contrast, so L=57L=57 and the circuit are fixed and only the moduli move, against an exact Pauli expansion of 159,338,497 terms at m=12m=12 for the microstructure of Figure 1 (§ 2.2).

The Block-Encoding Circuit

Registers

The system register carries 2m+⌈log⁡2nd⌉2m+\lceil\log_2 n_d\rceil qubits (nd=1n_d=1 for Poisson and nd=2n_d=2 for elasticity), ordered as xx, yy, then the dof qubit. The ancilla carries one shift-control qubit ss, the m−1m-1 clean carry ancillas aa of the incrementer and finally, the PREP register of nprepn_{\mathrm{prep}} qubits. Table 4 collects the resulting cost in closed form.

cellLLsystemcarryPREPtotalToffoli gates
1D Poisson3mmm−1m-122m+22m+24m4m
2D Poisson92m2mm−1m-143m+43m+48m8m
2D Poissonχ^{\chi}252m2mm−1m-173m+73m+716m−716m-7
2D elasticity172m+12m+1m−1m-163m+73m+78m+48m+4
2D elasticityχ^{\chi}572m+12m+1m−1m-193m+103m+1016m−116m-1

Table 4. Cost of the five encodings as a function of mm, exact at every m≥2m \ge2 and verified against measurement at every mm from 2 to 12. A superscript χ\chi marks a two-phase cell. LL is the number of LCU terms, the four middle columns are qubit counts, and the last is a gate count. Every cell also carries one shift control qubit, included in the total; it and the m−1m-1 carry ancillas are clean and shared by every shift in the circuit, so both are counted once. nprepn_{\mathrm{prep}} can exceed ⌈log⁡2L⌉\lceil\log_{2} L \rceil because the index is factored; see § 8.3. The Toffoli count is taken before transpilation; for the transpiled two-qubit count see Figure 4.

Only two entries depend on the mesh: the system register, at 2m+⌈log⁡2nd⌉2m+\lceil\log_2 n_d\rceil qubits, and the m−1m-1 carry ancillas, which are returned to ∣0⟩\lvert0\rangle and reused by every shift. Everything the decomposition controls is constant in mm: the term count, the PREP width, and the coefficients themselves. The Toffoli count is affine in mm at a slope of 8 per shifted register, which is the four controlled shifts of a homogeneous cell and the eight of a two-phase cell at 2m−22m-2 Toffoli each.

The Cyclic Shift

The shift is the controlled ripple-carry cyclic increment of § 3.1, drawn in Figure 3: 2m−22m-2 Toffoli, mm CNOT and m−1m-1 clean ancillas, transpiling to 11m−1011m-10 two-qubit gates.

Image not extracted

Figure 3. The controlled cyclic increment at m=4m = 4, the only primitive whose cost grows with the mesh.

PREP

Two different preparations are used,

PL∣0⟩=∑k∣ck∣α∣k⟩,\mathsf{P}_L\lvert0\rangle=\sum_k\sqrt{\frac{\lvert c_k\rvert}{\alpha}}\lvert k\rangle,
PR∣0⟩=∑ksgn⁡(ck)∣ck∣α∣k⟩,(48)\mathsf{P}_R\lvert0\rangle=\sum_k\operatorname{sgn}(c_k)\sqrt{\frac{\lvert c_k\rvert}{\alpha}}\lvert k\rangle, \tag*{(48)}

so that

U=(PR†⊗I)⋅SEL⋅(PL⊗I)(49)\mathcal{U}=(\mathsf{P}_R^\dagger\otimes I)\cdot\mathrm{SEL}\cdot(\mathsf{P}_L\otimes I) \tag*{(49)}

has top-left block ∑kckUk/α=K/α\sum_k c_k U_k/\alpha= \mathbf{K}/\alpha and no signs appear inside SELECT.

A standard dense PREP uses ⌈log⁡2L⌉\lceil\log_{2} L \rceil qubits. SELECT must then apply each UkU_k controlled on all nprepn_{\mathrm{prep}} qubits, which requires LL controlled blocks and O(L)O(L) Toffoli gates [9]. Every unitary in Equations (21), (39) and (42) is, however, a tensor product with one factor per register: a shift on xx, a shift on yy, a Pauli σr\sigma_r on the dof register, and at most one reflection R(a,b)R_{(a,b)}. Each factor is drawn from a small fixed set, three shifts per spatial register and four Pauli operators, so the terms are the combinations of those sets while the operations the circuit has to realize are their union. We therefore use a factored encoding of the index [23], in which SELECT applies one multiplexer per register and the choice made on one register is independent of the choice made on the others; for a Pauli LCU, factored encoding over qubits is standard. Consequently, SELECT cost is independent of LL; the Poisson operators at L=5L = 5 (FD) and L=9L = 9 (FE) compile to identical SELECT circuits. The small price to pay is a wider PREP register. The index kk is the concatenation of one field per factor: two bits for each shift exponent, two bits for σr\sigma_r, and, for a two-phase operator, three one-bit fields for the reflection: a flag gg that applies RχR_{\chi}, and bits aa and bb that select the offset in Equation (33). Unused codes receive zero amplitude. The register grows from ⌈log⁡2L⌉=4,5,5,6\lceil\log_{2} L \rceil= 4, 5, 5, 6 to nprep=4,7,6,9n_{\mathrm{prep}} = 4, 7, 6, 9 for the four cells (Table 4).

SELECT

SELECT applies one controlled operation for each value of each PREP field. The stages, in circuit order, are (see Appendix § 15):

  1. Spatial: on each of xx and yy, a controlled increment ScS_c and a controlled decrement Sc†S_c^{\dagger}, four controlled shifts in all.

  1. Dof: a controlled ZZ and XX, and in the two-phase elasticity cell a controlled iYiY, implemented as Ry(−π)R_y(-\pi). This stage is absent in the scalar cells.

  1. Reflection: in the two-phase cells, decrements on xx and yy controlled on aa and bb, RxR_x controlled on gg, and the matching increments. This is R(a,b)R_{(a,b)} of Equation (33), and adds four controlled shifts and one oracle query.

SELECT therefore contains four controlled shifts in a homogeneous cell and eight in a two-phase cell, at most three controlled single-qubit gates, and at most one oracle query, for every LL. Each controlled shift costs 2m−22m-2 Toffoli and each two-bit stage adds two, which gives the Toffoli counts of Table 4. Since SELECT does not depend on LL, each elasticity circuit exceeds its scalar counterpart by a constant number of gates.

Finally, every field is at most two bits wide, so each control line is computed by one Toffoli, and no gate in the four circuits has more than two controls. This matters for the following reason.

In Qiskit 2.4.2 [42], the transpiler synthesizes a multi-controlled XX with four or more controls using idle circuit qubits as clean ancillas, since a Qiskit circuit assumes every qubit starts in ∣0⟩\lvert0\rangle. A block encoding must act correctly on an arbitrary system state, so the transpiled circuit applies a different operator. The circuits here avoid this: the factored index of § 8.3 limits every gate to two controls, and we synthesize every multi-controlled gate explicitly. Appendix § 15 draws all four encodings as built.

Circuit Cost

Section 2.1 noted that the arbitrary-angle rotations of PREP dominate the fault-tolerant cost, each costing O(log⁡1/εsyn) T\mathcal{O}(\log1/\varepsilon_{\mathrm{syn}})\,T gates while everything else is Clifford and Toffoli. Table 5 splits the two-qubit count accordingly: PREP is constant in mm and SELECT grows.

cellnprepn_{\mathrm{prep}}slotsLLrot.PREP CXshare
2D scalar, homog.416929224%
2D scalar, 2-phase71282525324018%
2D elast., homog.6641712511416%
2D elast., 2-phase9512571021100446%

Table 5. Cost of the two preparations at m=12m = 12, against the whole circuit. Slots is 2nprep2^{n_{\mathrm{prep}}}, the addressable indices; LL of them carry amplitude and the rest are prepared at zero.

Figure 4 gives the transpiled two-qubit count from m=2m = 2 to 12; the slopes are 48 gates per qubit of resolution in the homogeneous cells and 96 in the two-phase cells. The homogeneous elasticity circuit costs a constant more than the homogeneous scalar one. The intercepts are 25 and 143 in the homogeneous cells and 209 and 1012 in the two-phase cells.

Image not extracted

Figure 4. Two-qubit gate count against mm for the four 2D cells, measured rather than fitted, for the circuit transpiled to {CX,U}\{\mathrm{CX}, U\} at optimization level 2 (Qiskit 2.4.2).

Tightness

The ratio α/∥K∥2\alpha/\lVert K\rVert_2 factors as

α∥K∥2=α∥K∥∞⏟LCU slack×∥K∥∞∥K∥2⏟norm gap.(50)\frac{\alpha}{\lVert K\rVert_2} = \underbrace{\frac{\alpha}{\lVert K\rVert_\infty}}_{\text{LCU slack}} \times\underbrace{\frac{\lVert K\rVert_\infty}{\lVert K\rVert_2}}_{\text{norm gap}} . \tag*{(50)}

Table 6 evaluates both factors. Both scalar cells have an LCU slack of 1 here: the four element contributions at a node share a sign, so α=∥K∥∞\alpha= \lVert K\rVert_\infty whenever some node lies entirely in the stiffer phase, as shown in § 6; for a microstructure with no such node, the checkerboard of that section for instance, the slack exceeds 1. The elasticity cells have a slack above 1 at every microstructure because contributions of opposite sign cancel (§ 7). For homogeneous elasticity, α/∥K∥2=1.1935×1.1625=1.3875\alpha/\lVert K\rVert_2 = 1.1935 \times1.1625 = 1.3875 at every mm; in the two-phase cells the norm gap, and with it the product, decreases as mm increases.

cellLCU slacknorm gapα/∥K∥2\alpha/\lVert K\rVert_2psuccp_{\mathrm{succ}}
2D scalar, homog.1.00001.33331.33330.563
2D scalar, 2-phase1.00001.34851.34850.550
2D elast., homog.1.19351.16251.38750.519
2D elast., 2-phase1.24731.17781.46910.463

Table 6. The two factors of Equation (50) at m=5m = 5, ν=0.3\nu= 0.3, E1/E2=3E_1/E_2 = 3. The LCU slack is independent of mm in all four cells; the norm gap is independent of mm only in the homogeneous cells. psuccp_{\mathrm{succ}} is the post-selection probability on the maximizing singular vector.

Implementation

This work is implemented in PyBlockEncode, an open-source Python library built on Qiskit 2.4.2 [42] and available at https://github.com/UW-ERSL/PyBlockEncode; the importable module is pyblockencode. The supported operators are

  • POISSON1D()

  • POISSON2D('fd')

  • POISSON2D('fe')

  • ELASTICITY2D(nu, E)

  • POISSON2D_2PHASE(vf, E1, E2)

  • ELASTICITY2D_2PHASE(nu, vf, E1, E2)

The size is given either as NN, the grid points per direction, or as m=log⁡2Nm = \log_{2} N qubits per direction; the encoded operator carries NN, N2N^{2} or 2N22N^{2} rows according to the cell. Listing 1 encodes the homogeneous elasticity cell at m=12m = 12.

Listing 1: The homogeneous elasticity cell at m=12m = 12.

from pyblockencode import blockencode, ELASTICITY2D

cell = ELASTICITY2D(nu=0.3)
qc, info = blockencode(cell, N=4096)

print(f"operator = {info.kind}")
print(f"dofs = {info.dofs:,} L = {info.L}")
print(f"alpha = {info.alpha:.4f}")
print(f"qubits = {info.qubits}")
print(f"Toffoli = {info.toffoli} CX = {info.cx}")
----- output ------------------------------
operator = elasticity2d
dofs = 33,554,432 L = 17
alpha = 6.0989
qubits = 43
Toffoli = 100 CX = 719

The two-phase operators place a centered square inclusion of volume fraction vf; every result reported here uses vf=1/4v_f = 1/4, for which the oracle is Equation (30). Listing 2 encodes the two-phase elasticity cell at m=12m = 12.

Listing 2: The two-phase elasticity cell at m=12m = 12.

from pyblockencode import blockencode, ELASTICITY2D_2PHASE

cell = ELASTICITY2D_2PHASE(nu=0.3, vf=0.25, E1=3, E2=1)
qc, info = blockencode(cell, N=4096)

print(f"operator = {info.kind}")
print(f"dofs = {info.dofs:,} L = {info.L}")
print(f"alpha = {info.alpha:.4f}")
print(f"qubits = {info.qubits}")
print(f"Toffoli = {info.toffoli} CX = {info.cx}")
----- output ------------------------------
operator = elasticity2d_2phase
dofs = 33,554,432 L = 57
alpha = 19.1209
qubits = 46
Toffoli = 191 CX = 2164

Conclusions

In this paper, we presented exact block encodings for 2D homogeneous elasticity, two-phase 2D Poisson, and two-phase 2D elasticity operators on periodic finite element grids, built on shift decomposition and a material reflection oracle. For homogeneous 2D elasticity at m=12m = 12 qubits per direction (33.6 million degrees of freedom), the LCU has 17 terms, against 18.9 million for the Pauli expansion.

Two-phase microstructures enter through one material reflection oracle conjugated by the four node-to-element shifts. The resulting LCUs have L=25L = 25 terms for scalar Poisson and L=57L = 57 for elasticity (4949 at ν=1/3\nu= 1/3), independent of the grid resolution NN, the volume fraction vfv_f, and the contrast r=E1/E2r = E_1/E_2; the subnormalizations are closed forms in the two moduli, independent of NN and vfv_f. The formulation is restricted to regular grids, periodic boundary conditions, and isotropic phases.

The periodic cell is the setting of computational homogenization that this paper targets. Non-periodic boundaries and unstructured meshes require additional de-periodization and mapping [17], which are not treated here.

Several extensions follow the same construction: (1) Arbitrary microstructures require only a different RχR_\chi, since neither LL nor αχ\alpha^\chi depends on χ\chi, so the open cost is the oracle itself: four CNOT and one controlled ZZ for a predicate with a closed form, and a table lookup for one without. (2) A functionally graded phase replaces the binary indicator with a discretized modulus, which the same split encodes as one reflection per bit plane, so LL grows with the bit depth rather than with the number of distinct moduli. (3) The decomposition applies unchanged to any translation-invariant finite element operator on a regular periodic grid, including the mass, convection-diffusion and Helmholtz operators, for which only the element matrix and hence the coefficients change. (4) Extending the work from 2D to 3D leave the node-to-element map and the reflection unchanged, with eight incident elements in place of four, and the scalar cell follows with new constants and no new argument, while the vector cell requires padding the 3×33 \times3 nodal blocks into a two-qubit dof register, which is where the additional terms and subnormalization arise [29]. (5) A more efficient PREP [43] may be desirable, especially in 3D.

Two directions lie outside the present formulation. Nonlinear elasticity re-encodes, at every Newton step, a tangent operator whose element moduli vary with the current state; whether the oracle can be updated rather than rebuilt at each iteration is open. Finally, the end-to-end cost of a QSVT-based solve [4] remains to be quantified, and the subnormalizations reported here need to be quantified.

Acknowledgments

The author acknowledges the Vilas Associate Grant from the University of Wisconsin Graduate School.

Declarations

The author declares no conflict of interest.

Code availability

The code developed in this work is available at https://github.com/UW-ERSL/PyBlockEncode.

Use of AI Tools

The author used generative AI assistants, specifically, Claude (Anthropic) and Gemini (Google) during the preparation of this manuscript. The tools were used for code development and drafting of text. All content was reviewed, verified, and edited by the author, who takes full responsibility for the accuracy and integrity of the work.

The Four Circuits

Figures A.1 to A.4 give the four encodings at m=2m = 2, with only the two preparations boxed and every other gate drawn as built. The gaps marked reindex are where the shift-control qubit ss is cleared, the PREP index is recoded for the next term, and ss is set again.

Image not extracted

Figure A.1. Scalar homogeneous cell at m=2m = 2, L=9L = 9, α=16/3\alpha= 16/3, on 3m+4=103m + 4 = 10 qubits. The four bracketed stages SxS_x, Sx†S_x^\dagger, SyS_y, Sy†S_y^\dagger are the two multiplexers of § 8.4, one per spatial register. Within each, the Toffoli before and after computes and uncomputes the SELECT control from a two-bit field, and the XX gates around it choose which value of that field the shift controls.

Image not extracted

Figure A.2. Scalar two-phase cell at m=2m = 2, L=25L = 25, α=16\alpha= 16 at E1/E2=3E_1/E_2 = 3, on 3m+7=133m + 7 = 13 qubits. The first half is Figure A.1; the three brackets SabS_{ab}, RχR_\chi, Sab†S_{ab}^\dagger at the right are the conjugation of Equation (33) read right to left, a conditional shift on each spatial register, the oracle controlled on gg, and the shifts undone.

Image not extracted

Figure A.3. Elasticity homogeneous cell at m=2m = 2, L=17L = 17, α=6.099\alpha= 6.099 at ν=0.3\nu= 0.3, on 3m+7=133m + 7 = 13 qubits.

Image not extracted

Figure A.4. Elasticity two-phase cell at m=2m = 2, L=57L = 57, α=19.121\alpha= 19.121 at ν=0.3\nu= 0.3 and E1/E2=3E_1/E_2 = 3, on 3m+10=163m + 10 = 16 qubits. Every stage of the construction appears: the four spatial multiplexers, the dof multiplexer ZZ, XX, iYiY, and the reflection stage, whose three brackets SabS_{ab}, RxR_x, Sab†S_{ab}^{\dagger} are the conjugation of Equation (33) read right to left.

CRediT authorship contribution statement

Krishnan Suresh: Conceptualization, Methodology, Software, Validation, Writing – original draft.

References

  1. [1]Aram W Harrow, Avinatan Hassidim, and Seth Lloyd. Quantum algorithm for linear systems of equations. Physical review letters, 103(15):150502, 2009.pubmed.ncbi.nlm.nih.gov/19905613
  2. [2]Andrew M Childs, Robin Kothari, and Rolando D Somma. Quantum algorithm for systems of linear equations with exponentially improved dependence on precision. SIAM Journal on Computing, 46(6):1920–1950, 2017.
  3. [3]Pedro CS Costa, Dong An, Yuval R Sanders, Yuan Su, Ryan Babbush, and Dominic W Berry. Optimal scaling quantum linear-systems solver via discrete adiabatic theorem. PRX Quantum, 3(4):040303, 2022.arxiv.org/abs/2111.08152
  4. [4]András Gilyén, Yuan Su, Guang Hao Low, and Isaac L Chuang. Quantum singular value transformation and beyond: exponential improvements for quantum matrix arithmetics. In Proceedings of the 51st Annual ACM SIGACT Symposium on Theory of Computing (STOC), pages 193–204, 2019. arXiv:1806.01838.DOI
  5. [5]Dominic W Berry, Andrew M Childs, Richard Cleve, Robin Kothari, and Rolando D Somma. Simulating Hamiltonian dynamics with a truncated Taylor series. Physical review letters, 114(9):090502, 2015.pubmed.ncbi.nlm.nih.gov/25793789
  6. [6]Daan Camps and Roel Van Beumeen. FABLE: Fast approximate quantum circuits for block-encodings. 2022 IEEE International Conference on Quantum Computing and Engineering (QCE), pages 104–113, 2022. arXiv:2205.00081.DOI
  7. [7]Parker Kuklinski, Benjamin Rempfer, Justin Elenewski, and Kevin Obenland. Efficient block-encodings require structure, 2025. arXiv:2509.19667.arxiv.org/abs/2509.19667
  8. [8]Andrew M Childs and Nathan Wiebe. Hamiltonian simulation using linear combinations of unitary operations. Quantum Information & Computation, 12(11-12):901–924, 2012.
  9. [9]Ryan Babbush, Craig Gidney, Dominic W Berry, Nathan Wiebe, Jarrod McClean, Alexandru Paler, Austin Fowler, and Hartmut Neven. Encoding electronic spectra in quantum circuits with linear T complexity. Physical Review X, 8(4):041015, 2018.DOI
  10. [10]Krishnan Suresh. Applied Quantum Computing for Engineers: Linear Solvers and Optimization. Independently published, first edition, 2026. ISBN-13: 979-8244366846.
  11. [11]Lukas Hantzko, Lennart Binkowski, and Sabhyata Gupta. Tensorized Pauli decomposition algorithm. Physica Scripta, 99(8):085128, 2024.
  12. [12]Océane Koska, Marc Baoubin, and Arnaud Gazda. A tree-approach Pauli decomposition algorithm with application to quantum computing. In ISC High Performance 2024 Research Paper Proceedings (39th International Conference), pages 1–11. Prometeus GmbH, 2024.arxiv.org/abs/2403.11644
  13. [13]Rocco Monteiro Nunes Pesce and Paul D Stevenson. H2ZIXY: Pauli spin matrix decomposition of real symmetric matrices. arXiv preprint arXiv:2111.00627, 2021.arxiv.org/abs/2111.00627
  14. [14]Timothy N. Georges, Bjorn K. Berntson, Christoph Sünderhauf, and Aleksei V. Ivanov. Pauli decomposition via the fast Walsh–Hadamard transform. New Journal of Physics, 27(3):033004, 2025.DOI
  15. [15]Chao Lu, Muralikrishnan Gopalakrishnan Meena, Eduardo Antonio Coello Perez, Kalyana Chakravarthi Gottiparthi, and Seongmin Kim. Memory-, circuit-, and ansatz-efficient vqls for cfd on hybrid quantum-hpc systems. arXiv preprint arXiv:2608.09661, 2026.arxiv.org/abs/2608.09661
  16. [16]Yongchun Xu and Heng Hu. Decomposition-free variational quantum linear solver: Application in computational mechanics. Computer Methods in Applied Mechanics and Engineering, 447:118396, 2025.DOI
  17. [17]Tyler Kharazi, Ahmad M Alkadri, Jin-Peng Liu, Kranthi K Mandadapu, and K Birgitta Whaley. Explicit block encodings of boundary value problems for many-body elliptic operators. Quantum, 9:1764, 2025.DOI
  18. [18]Andreas Sturm and Niclas Schillo. Efficient and explicit block encoding of finite difference discretizations of the Laplacian. ACM Transactions on Quantum Computing, 2025.
  19. [19]Ahmad M Alkadri, Tyler D Kharazi, K Birgitta Whaley, and Kranthi K Mandadapu. A quantum algorithm for the finite element method. arXiv preprint arXiv:2510.18150, 2025.arxiv.org/abs/2510.18150
  20. [20]Thomas Hogancamp, Reuben Demirdjian, and Daniel Gunlycke. Linear-combination-of-unitaries decomposition for the Laplace operator. Physical Review A, 114(1):012455, 2026.DOI
  21. [21]Hai-Ling Liu, Yu-Sen Wu, Lin-Chun Wan, Shi-Jie Pan, Su-Juan Qin, Fei Gao, and Qiao-Yan Wen. Variational quantum algorithm for the Poisson equation. Physical Review A, 104(2):022418, 2021.DOI
  22. [22]Abevnya Gnanasekaran and Amit Surana. Efficient quantum access model for sparse structured matrices using linear combination of “things”. Physical Review A, 113(2):022437, 2026.
  23. [23]Daan Camps, Lin Lin, Roel Van Beumeen, and Chao Yang. Explicit quantum circuits for block encodings of certain sparse matrices. SIAM Journal on Matrix Analysis and Applications, 45(1):801–827, 2024. arXiv:2203.10236.arxiv.org/abs/2203.10236
  24. [24]Christoph Sünderhauf, Earl Campbell, and Joan Camps. Block-encoding structured matrices for data input in quantum computing. Quantum, 8:1226, 2024. arXiv:2302.10949.DOI
  25. [25]Sven Danz, Tobias Stollenwerk, and Alessandro Ciani. An implementation of quantum oracles for the finite element method. SIAM Journal on Scientific Computing, 48(3):B386–B419, June 2026.DOI
  26. [26]Abhishek Arora, Benjamin M Ward, and Caglar Ozkay. An implementation of the finite element method in hybrid classical/quantum computers. Finite Elements in Analysis and Design, 248:104354, 2025.arxiv.org/abs/2411.09038
  27. [27]Leonhard Hölscher, Oliver Ahrend, Lukas Karch, Carlotta L’Estocq, Marc Marfany Andreu, Tobias Stollenwerk, Frank K Wilhelm, and Julia Kowalski. End-to-end quantum algorithm for topology optimization in structural mechanics. Quantum Science and Technology, 11(2):025029, 2026.arxiv.org/abs/2510.07280
  28. [28]Matthias Deiml and Daniel Peterseim. Quantum realization of the finite element method. arXiv preprint arXiv:2403.19512, 2024.DOI
  29. [29]Austin Pechan, John Golden, and Daniel O’Malley. Block encoding of the three-dimensional heterogeneous Poisson equation with application to fracture flow. Physical Review Applied, 25(4):044038, 2026.DOI
  30. [30]Hiroshi Yano and Yuki Sato. Quantum framework for parameterizing partial differential equations via diagonal block-encoding. arXiv preprint arXiv:2603.01358, 2026.arxiv.org/abs/2603.01358
  31. [31]Jishnu Mahmud and Rebekah Herrman. Moment-structured block encodings of periodic finite-difference operators. arXiv preprint arXiv:2607.11596, 2026.arxiv.org/abs/2607.11596
  32. [32]Thomas G. Draper. Addition on a quantum computer. arXiv preprint quant-ph/0008033, 2000.DOI
  33. [33]Craig Gidney. Halving the cost of quantum addition. Quantum, 2:74, 2018.DOI
  34. [34]Steven A. Cuccaro, Thomas G. Draper, Samuel A. Kutin, and David Petrie Moulton. A new quantum ripple-carry addition circuit. arXiv preprint quant-ph/0410184, 2004.DOI
  35. [35]Michel O. Deville, Paul F Fischer, and Ernest H Mund. High-order methods for incompressible fluid flow, volume 1. Cambridge university press Cambridge, 2002.
  36. [36]Charles F Van Loan. The ubiquitous Kronecker product. Journal of computational and applied mathematics, 123(1-2):85–100, 2000.DOI
  37. [37]Olgierd Cecil Zienkiewicz, Robert Leroy Taylor, and Jian Z Zhu. The finite element method: its basis and fundamentals. Elsevier, 2013.
  38. [38]Felix Givois, Matthias Kabel, and Nicolas Gauger. QFT-based homogenization. arXiv preprint arXiv:2207.12949, 2022.arxiv.org/abs/2207.12949
  39. [39]Burigede Liu, Michael Ortiz, and Fehmi Cirak. Towards quantum computational mechanics. Computer Methods in Applied Mechanics and Engineering, 432:117403, 2024.
  40. [40]Yiren Wang, Michael Ortiz, and Fehmi Cirak. QAFE2: Quantum accelerated multiscale finite element analysis. Computer Methods in Applied Mechanics and Engineering, 461:119204, 2026.DOI
  41. [41]Loïc Balazi, Matthias Deiml, and Daniel Peterseim. Quantum enhanced numerical homogenization. arXiv preprint arXiv:2603.28521, 2026.arxiv.org/abs/2603.28521
  42. [42]Ali Javadi-Abhari, Matthew Treinish, Kevin Krsulich, Christopher J Wood, Jake Lishman, Julien Gacon, Simon Martiel, Paul D Nation, Lev S Bishop, Andrew W Cross, et al. Quantum computing with qiskit. arXiv preprint arXiv:2405.08810, 2024.arxiv.org/abs/2405.08810
  43. [43]Krishnan Suresh and Sanjay Suresh. Pyencode: an open-source library for structured quantum state preparation. Quantum Science and Technology, 2026.arxiv.org/abs/2603.28259

Paper details

vlnc:W_k93sf9ez
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 8

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Computer Science

Claude (Anthropic), Gemini (multiple versions)

View statement
Looking for related papers…
Contents