Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2609.33686

Simultaneously Query-Optimal Quantum Linear-System Algorithm

Minbo Gao, Zhengfeng Ji, Chenghua Liu

quant-phSubmitted Sep 27, 2026• Announced Sep 29, 2026• 1 citation
arXivHTMLPDF

Abstract

Quantum linear-system algorithms access the coefficient matrix through a block-encoding oracle $U_A$ and the right-hand side through a state-preparation oracle $U_b$. The optimal query complexities of these two oracles are known separately: $O(κ\log(1/\varepsilon))$ queries to $U_A$ and $O(κ/s)$ queries to $U_b$, where $s=α\|A^{-1}b\|$. Whether a single algorithm can achieve both bounds simultaneously has remained open. We resolve this question affirmatively. Given a constant-factor estimate of $s$, we give a quantum linear-system algorithm that prepares the normalized solution to error $\varepsilon$ using simultaneously $ O\left(κ\log\frac1\varepsilon\right)$ queries to $U_A$ and $O\left(κ/{s}\right)$ queries to $U_b$. The algorithm separates state preparation from precision refinement, so only the matrix-query count depends on the target accuracy. We also prove a matching same-instance worst-case lower bound: for every quantum linear-system algorithm and every solution-norm scale, there exists a single Hermitian instance on which that algorithm requires both $Ω(κ\log(1/\varepsilon))$ matrix queries and $Ω(κ/s)$ state-preparation queries. Thus the two oracle complexities can be optimized simultaneously, and the resulting bounds are jointly optimal.

AI Summary

Quantum linear-system algorithms access $A$ through a block-encoding oracle $U_A$ and $b$ through a preparation oracle $U_b$. This paper asks whether one algorithm can attain the separately known optimal query bounds for both oracles, rather than trading matrix queries against repeated preparation of the right-hand side. Given an exact $(\alpha,a,0)$ block-encoding, $\kappa\ge\alpha\|A^{-1}\|$, and a constant-factor estimate of $s=\alpha\|A^{-1}b\|$, the algorithm prepares the normalized solution to error $\varepsilon$ with success probability at least $2/3$, using simultaneously $O(\kappa\log(1/\varepsilon))$ matrix queries and $O(\kappa/s)$ preparation queries. Additional gates scale as $O(\kappa(a+1)\log(1/\varepsilon)+\kappa(1+\log d)/s)$. The construction uses a transducer to prepare a constant-accuracy state with separately controlled oracle costs. A real Krylov-space invariant ensures that its projection onto an auxiliary kernel has exactly the required direction. Polynomial kernel filtering and a correction transformation then refine precision using only matrix queries. A computation-history reduction from parity, combined with a pointwise hybrid argument for preparation queries, establishes both matching lower bounds on a single Hermitian instance at every promised solution-norm scale. The upper bound assumes the supplied norm estimate and access to controlled oracle calls and adjoints. The explicit implementation has large constant overheads. For a fixed approximate block-encoding with error $\delta_A$, the same asymptotic costs yield solution error at most $\varepsilon+2\kappa\delta_A/\alpha$, provided $\kappa\delta_A/\alpha\le1/4$; this robustness statement concerns a fixed encoded matrix, not time-varying oracle noise.

Little sign of AI-generated text

Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.

Comments
37 pages

Related papers

A Near-Optimal Joint Lower Bound for Sparse Quantum Linear System Solvers

Dhrumil Patel

Quantum linear system solvers form one of the central algorithmic primitives in quantum computing, with applications ranging from differential equations and optimization to machine learning. Their cost is commonly measured through query complexity, which counts the number of oracle calls needed to access the input matrix. In the sparse-access model, this complexity is governed by three parameters: the condition number $κ$, the sparsity $s$ of the input matrix, and the target precision $\varepsilon$. The dependence on $κ$ and $\varepsilon$ is already well understood through the lower bound $Ω(κ\log(1/\varepsilon))$, which matches the best known scaling in these parameters. Once the sparsity $s$ is included, however, the expected lower bound has long been conjectured to be $Ω(κ\sqrt{s}\log(1/\varepsilon))$. Recent work by Mori et al. [Quantum Sci. Tech. 11 035063 (2026)] made an important step towards this goal by establishing the lower bound $Ω(κ\sqrt{s})$ for constant error $\varepsilon$. In this work, we complete the picture and prove the full joint lower bound $Ω(κ\sqrt{s}\log(1/\varepsilon))$ in the sparse-access model, thereby establishing the anticipated dependence on all three parameters simultaneously.

quant-pharXiv:2609.22445•Sep 22, 2026•5 citations

Quantum linear system algorithm with optimal queries to initial state preparation

Guang Hao Low, Yuan Su

Quantum algorithms for linear systems produce the solution state A − 1 | b ⟩ by querying two oracles: O A that block encodes the coefficient matrix and O b that prepares the initial state. We present a quantum linear system algorithm making Θ ( 1 / p ) queries to O b , which is optimal in the success probability, and O ( κ log ⁡ ( 1 / p ) ( log ⁡ log ⁡ ( 1 / p ) + log ⁡ ( 1 / ϵ ) ) ) queries to O A , nearly optimal in all parameters including the condition number and accuracy. Notably, our complexity scaling of initial state preparation holds even when p is not known a priori . This contrasts with recent results achieving O ( κ log ⁡ ( 1 / ϵ ) ) complexity to both oracles, which, while optimal in O A , is highly suboptimal in O b as κ can be arbitrarily larger than 1 / p . In various applications such as solving differential equations, preparing ground states of operators with real spectra, and estimating and transforming eigenvalues of non-normal matrices, we can further improve the dependence on p using a block preconditioning scheme to nearly match or outperform best p

arXiv:2410.18178•Quantum•Mar 23, 2026•41 citations

Linear combination of Schrödingerization for quantum linear systems with optimal matrix-query complexity

Yin Yang, Yue Yu, Long Zhang

Quantum linear systems algorithms (QLSAs) aim to solve linear systems $A\bb{x}=\bb{b}$ exponentially faster than classical methods under certain conditions. In this work, we develop quantum algorithms for solving linear algebraic equations from an ODE-based perspective. Inspired by the linear combination of Hamiltonian simulation (LCHS) representation in the Fourier approach \cite{Childs2017QLSA}, we express the solution $\bb{x}$ as a linear combination of solutions to a system of linear convection equations, which become Schrödinger-type equations with unitary evolutions in the Fourier domain. We refer to this representation as LC-Schrödingerization. Based on this result, we construct an LCHS-based quantum algorithm with two LCHS instances: one for time-marching and one for numerical integration. The key construction uses the derivative of a Gaussian-smoothed hat function and recovers the solution over a fixed auxiliary interval. This permits a truncation time independent of the target accuracy and avoids the loss in success probability from selecting a single grid point. Periodization and explicit Fourier coefficients provide the corresponding projection error bounds. Under the stated oracle assumptions and given a constant-factor estimate of the solution norm, direct simulation of the select operators and block preconditioning achieve the optimal matrix-query complexity $\mathcal{O}(κ_A\log\frac1\varepsilon)$ without using variable-time amplitude amplification (VTAA). The same upper bound holds for queries to the right-hand-side preparation oracle.

quant-pharXiv:2609.19676•Sep 18, 2026•2 citations

Faster quantum linear system solver beyond the condition number

Alexander M. Dalzell, Jianqiang Li, Yuan Su

The spectral condition number is a widely adopted measure of worst-case cost for quantum linear system solvers. Yet it can significantly overestimate the actual runtime for a typical problem instance. We present two quantum algorithms that produce the normalized solution $|x\rangle$ of linear system $Ax=| b \rangle$ to accuracy $ε$ with complexity independent of the condition number $κ=\lVert A^{-1}\rVert$. We focus on the standard input model where $A$ is accessed through a block encoding and $| b \rangle$ is prepared by a unitary. But we also introduce an affine dilation model that encodes $A$ and $| b \rangle$ jointly, allowing further refinements of the query complexity. Our truncation-based solver makes an optimal number of queries to $| b \rangle$ and $\operatorname{\mathbf{O}}\left(κ_{\mathrm{eff}}\operatorname{polylog}\left(\frac{κ_{\mathrm{eff}}}ε\right)\right)$ queries to $A$. We prove a family of upper bounds on the effective condition number, including $κ_{\mathrm{eff}}\leq\frac{\lVert(A^\dagger A)^{-t/2}|x\rangle\rVert^{1/t}}{ε^{1/t}}$ for positive even integer $t$ and $κ_{\mathrm{eff}}\leq\frac{\lVert A^{-1\dagger}(A^\dagger A)^{-(t-1)/2}|x\rangle\rVert^{1/t}}{ε^{1/t}}$ for positive odd $t$, overcoming the $κ$-barrier. Our filtering-based solver is extremely simple with a favorable runtime prefactor. In particular, the solver has query complexity $3\frac{\lVert A^{-1\dagger}|x\rangle\rVert}ε$ to leading order when the solution norm is known. We then present a similarly simple solution norm estimator with the same asymptotic cost up to logarithmic factors. Our quantum linear system solvers thus substantially improve a recent algorithm of Li, enabling faster quantum linear system solving beyond the condition number.

quant-pharXiv:2607.07691•Jul 9, 2026•9 citations

Optimal Query Complexity for Ground-State Preparation

Boyang Chen, Minbo Gao, Xinzhao Wang +1 more

We determine the optimal query complexity of ground-state preparation to trace-distance error $\varepsilon$ when an energy threshold in the spectral gap is known. Let $U_H$ be an $α$-block-encoding of a Hamiltonian with unique ground state $|ψ_0\rangle$, and suppose $|\langleψ_0|U_I|0\rangle|\geγ$ for a state-preparation oracle $U_I$. The threshold lies at least $Δ/2$ above the ground-state energy and at least $Δ/2$ below every excited-state energy. We give two algorithms that prepare a state within trace distance $\varepsilon$ of the ground state. One uses $O((α/Δ)(γ^{-1}+\log(1/\varepsilon)))$ calls to $U_H$ in expectation; the other uses $O((α/(γΔ))\log(1/\varepsilon))$ calls to $U_H$ in the worst case. We prove a lower bound matching the expected query count; the corresponding worst-case lower bound follows from Somma and de Wolf [SdW26]. The respective bounds on calls to $U_I$ are $O(1/γ)$ in expectation and $O(γ^{-1}\log(1/\varepsilon))$ in the worst case. On $(N+1)$-dimensional systems, these $U_I$ bounds are also optimal when the expected or worst-case count of $U_H$ calls, respectively, is $o((α/Δ)\sqrt N)$. Both algorithms use a constant-accuracy spectral filter to construct a purifier, which we then sequentially compose during amplitude amplification to prepare a state with constant overlap with the ground state. The expected-query algorithm repeats the preparation followed by one high-accuracy spectral filter until success. The worst-case algorithm uses filters of increasing accuracy and limits the total number of queries.

quant-pharXiv:2609.35668•Sep 29, 2026•1 citation

An Optimal Quantum Linear Systems Algorithm

Carlos Bravo-Prieto, Aram W. Harrow, Robin Kothari

In the quantum linear systems problem (QLSP), we are given query access to a $d$-sparse $N\times N$ matrix $A$ with condition number $κ$, and the ability to prepare a quantum state proportional to a vector $\vec b$. The goal is to output an $ε$-approximation to the quantum state proportional to the solution $\vec x$ of $A\vec{x}=\vec{b}$. Following a long line of work, the best previously known quantum algorithms for the QLSP had query complexities $O(κd\log(1/ε))$ and $κ\sqrt d(κd/ε)^{o(1)}$, while the best known lower bounds were $Ω(κ\log(1/ε))$ and $Ω(κ\sqrt d)$. We improve these bounds and show that the complexity of the QLSP is $Θ(κ\sqrt d\log(1/ε))$. We also resolve an open problem of Berry and Childs by showing that any $N\times N$ unitary can be implemented with bounded error using $O(\sqrt N)$ queries to its matrix entries.

quant-pharXiv:2609.35660•Sep 29, 2026•9 citations
Quantum linear-system algorithms access the coefficient matrix through a block-encoding oracle UAU_AUA​ and the right-hand side through a state-preparation oracle UbU_bUb​. The optimal query complexities of these two oracles are known separately: O(κlog⁡(1/ε))O(κ\log(1/\varepsilon))O(κlog(1/ε)) queries to UAU_AUA​ and O(κ/s)O(κ/s)O(κ/s) queries to UbU_bUb​, where s=α∥A−1b∥s=α\|A^{-1}b\|s=α∥A−1b∥. Whether a single algorithm can achieve both bounds simultaneously has remained open.
We resolve this question affirmatively. Given a constant-factor estimate of sss, we give a quantum linear-system algorithm that prepares the normalized solution to error ε\varepsilonε using simultaneously O(κlog⁡1ε) O\left(κ\log\frac1\varepsilon\right)O(κlogε1​) queries to UAU_AUA​ and O(κ/s)O\left(κ/{s}\right)O(κ/s) queries to UbU_bUb​. The algorithm separates state preparation from precision refinement, so only the matrix-query count depends on the target accuracy.
We also prove a matching same-instance worst-case lower bound: for every quantum linear-system algorithm and every solution-norm scale, there exists a single Hermitian instance on which that algorithm requires both Ω(κlog⁡(1/ε))Ω(κ\log(1/\varepsilon))Ω(κlog(1/ε)) matrix queries and Ω(κ/s)Ω(κ/s)Ω(κ/s) state-preparation queries. Thus the two oracle complexities can be optimized simultaneously, and the resulting bounds are jointly optimal.
Quantum linear-system algorithms access AAA through a block-encoding oracle UAU_AUA​ and bbb through a preparation oracle UbU_bUb​. This paper asks whether one algorithm can attain the separately known optimal query bounds for both oracles, rather than trading matrix queries against repeated preparation of the right-hand side.
Given an exact (α,a,0)(\alpha,a,0)(α,a,0) block-encoding, κ≥α∥A−1∥\kappa\ge\alpha\|A^{-1}\|κ≥α∥A−1∥, and a constant-factor estimate of s=α∥A−1b∥s=\alpha\|A^{-1}b\|s=α∥A−1b∥, the algorithm prepares the normalized solution to error ε\varepsilonε with success probability at least 2/32/32/3, using simultaneously O(κlog⁡(1/ε))O(\kappa\log(1/\varepsilon))O(κlog(1/ε)) matrix queries and O(κ/s)O(\kappa/s)O(κ/s) preparation queries. Additional gates scale as O(κ(a+1)log⁡(1/ε)+κ(1+log⁡d)/s)O(\kappa(a+1)\log(1/\varepsilon)+\kappa(1+\log d)/s)O(κ(a+1)log(1/ε)+κ(1+logd)/s).
The construction uses a transducer to prepare a constant-accuracy state with separately controlled oracle costs. A real Krylov-space invariant ensures that its projection onto an auxiliary kernel has exactly the required direction. Polynomial kernel filtering and a correction transformation then refine precision using only matrix queries. A computation-history reduction from parity, combined with a pointwise hybrid argument for preparation queries, establishes both matching lower bounds on a single Hermitian instance at every promised solution-norm scale.
The upper bound assumes the supplied norm estimate and access to controlled oracle calls and adjoints. The explicit implementation has large constant overheads. For a fixed approximate block-encoding with error δA\delta_AδA​, the same asymptotic costs yield solution error at most ε+2κδA/α\varepsilon+2\kappa\delta_A/\alphaε+2κδA​/α, provided κδA/α≤1/4\kappa\delta_A/\alpha\le1/4κδA​/α≤1/4; this robustness statement concerns a fixed encoded matrix, not time-varying oracle noise.
Quantum linear system solvers form one of the central algorithmic primitives in quantum computing, with applications ranging from differential equations and optimization to machine learning. Their cost is commonly measured through query complexity, which counts the number of oracle calls needed to access the input matrix. In the sparse-access model, this complexity is governed by three parameters: the condition number κκκ, the sparsity sss of the input matrix, and the target precision ε\varepsilonε. The dependence on κκκ and ε\varepsilonε is already well understood through the lower bound Ω(κlog⁡(1/ε))Ω(κ\log(1/\varepsilon))Ω(κlog(1/ε)), which matches the best known scaling in these parameters. Once the sparsity sss is included, however, the expected lower bound has long been conjectured to be Ω(κslog⁡(1/ε))Ω(κ\sqrt{s}\log(1/\varepsilon))Ω(κs​log(1/ε)). Recent work by Mori et al. [Quantum Sci. Tech. 11 035063 (2026)] made an important step towards this goal by establishing the lower bound Ω(κs)Ω(κ\sqrt{s})Ω(κs​) for constant error ε\varepsilonε. In this work, we complete the picture and prove the full joint lower bound Ω(κslog⁡(1/ε))Ω(κ\sqrt{s}\log(1/\varepsilon))Ω(κs​log(1/ε)) in the sparse-access model, thereby establishing the anticipated dependence on all three parameters simultaneously.
Quantum algorithms for linear systems produce the solution state A − 1 | b ⟩ by querying two oracles: O A that block encodes the coefficient matrix and O b that prepares the initial state. We present a quantum linear system algorithm making Θ ( 1 / p ) queries to O b , which is optimal in the success probability, and O ( κ log ⁡ ( 1 / p ) ( log ⁡ log ⁡ ( 1 / p ) + log ⁡ ( 1 / ϵ ) ) ) queries to O A , nearly optimal in all parameters including the condition number and accuracy. Notably, our complexity scaling of initial state preparation holds even when p is not known a priori . This contrasts with recent results achieving O ( κ log ⁡ ( 1 / ϵ ) ) complexity to both oracles, which, while optimal in O A , is highly suboptimal in O b as κ can be arbitrarily larger than 1 / p . In various applications such as solving differential equations, preparing ground states of operators with real spectra, and estimating and transforming eigenvalues of non-normal matrices, we can further improve the dependence on p using a block preconditioning scheme to nearly match or outperform best p
Quantum linear systems algorithms (QLSAs) aim to solve linear systems $A\bb{x}=\bb{b}$ exponentially faster than classical methods under certain conditions. In this work, we develop quantum algorithms for solving linear algebraic equations from an ODE-based perspective. Inspired by the linear combination of Hamiltonian simulation (LCHS) representation in the Fourier approach \cite{Childs2017QLSA}, we express the solution $\bb{x}$ as a linear combination of solutions to a system of linear convection equations, which become Schrödinger-type equations with unitary evolutions in the Fourier domain. We refer to this representation as LC-Schrödingerization. Based on this result, we construct an LCHS-based quantum algorithm with two LCHS instances: one for time-marching and one for numerical integration. The key construction uses the derivative of a Gaussian-smoothed hat function and recovers the solution over a fixed auxiliary interval. This permits a truncation time independent of the target accuracy and avoids the loss in success probability from selecting a single grid point. Periodization and explicit Fourier coefficients provide the corresponding projection error bounds. Under the stated oracle assumptions and given a constant-factor estimate of the solution norm, direct simulation of the select operators and block preconditioning achieve the optimal matrix-query complexity O(κAlog⁡1ε)\mathcal{O}(κ_A\log\frac1\varepsilon)O(κA​logε1​) without using variable-time amplitude amplification (VTAA). The same upper bound holds for queries to the right-hand-side preparation oracle.
The spectral condition number is a widely adopted measure of worst-case cost for quantum linear system solvers. Yet it can significantly overestimate the actual runtime for a typical problem instance. We present two quantum algorithms that produce the normalized solution ∣x⟩|x\rangle∣x⟩ of linear system Ax=∣b⟩Ax=| b \rangleAx=∣b⟩ to accuracy εεε with complexity independent of the condition number κ=∥A−1∥κ=\lVert A^{-1}\rVertκ=∥A−1∥. We focus on the standard input model where AAA is accessed through a block encoding and ∣b⟩| b \rangle∣b⟩ is prepared by a unitary. But we also introduce an affine dilation model that encodes AAA and ∣b⟩| b \rangle∣b⟩ jointly, allowing further refinements of the query complexity. Our truncation-based solver makes an optimal number of queries to ∣b⟩| b \rangle∣b⟩ and O⁡(κeffpolylog⁡(κeffε))\operatorname{\mathbf{O}}\left(κ_{\mathrm{eff}}\operatorname{polylog}\left(\frac{κ_{\mathrm{eff}}}ε\right)\right)O(κeff​polylog(εκeff​​)) queries to AAA. We prove a family of upper bounds on the effective condition number, including κeff≤∥(A†A)−t/2∣x⟩∥1/tε1/tκ_{\mathrm{eff}}\leq\frac{\lVert(A^\dagger A)^{-t/2}|x\rangle\rVert^{1/t}}{ε^{1/t}}κeff​≤ε1/t∥(A†A)−t/2∣x⟩∥1/t​ for positive even integer ttt and κeff≤∥A−1†(A†A)−(t−1)/2∣x⟩∥1/tε1/tκ_{\mathrm{eff}}\leq\frac{\lVert A^{-1\dagger}(A^\dagger A)^{-(t-1)/2}|x\rangle\rVert^{1/t}}{ε^{1/t}}κeff​≤ε1/t∥A−1†(A†A)−(t−1)/2∣x⟩∥1/t​ for positive odd ttt, overcoming the κκκ-barrier. Our filtering-based solver is extremely simple with a favorable runtime prefactor. In particular, the solver has query complexity 3∥A−1†∣x⟩∥ε3\frac{\lVert A^{-1\dagger}|x\rangle\rVert}ε3ε∥A−1†∣x⟩∥​ to leading order when the solution norm is known. We then present a similarly simple solution norm estimator with the same asymptotic cost up to logarithmic factors. Our quantum linear system solvers thus substantially improve a recent algorithm of Li, enabling faster quantum linear system solving beyond the condition number.
We determine the optimal query complexity of ground-state preparation to trace-distance error ε\varepsilonε when an energy threshold in the spectral gap is known. Let UHU_HUH​ be an ααα-block-encoding of a Hamiltonian with unique ground state ∣ψ0⟩|ψ_0\rangle∣ψ0​⟩, and suppose ∣⟨ψ0∣UI∣0⟩∣≥γ|\langleψ_0|U_I|0\rangle|\geγ∣⟨ψ0​∣UI​∣0⟩∣≥γ for a state-preparation oracle UIU_IUI​. The threshold lies at least Δ/2Δ/2Δ/2 above the ground-state energy and at least Δ/2Δ/2Δ/2 below every excited-state energy. We give two algorithms that prepare a state within trace distance ε\varepsilonε of the ground state. One uses O((α/Δ)(γ−1+log⁡(1/ε)))O((α/Δ)(γ^{-1}+\log(1/\varepsilon)))O((α/Δ)(γ−1+log(1/ε))) calls to UHU_HUH​ in expectation; the other uses O((α/(γΔ))log⁡(1/ε))O((α/(γΔ))\log(1/\varepsilon))O((α/(γΔ))log(1/ε)) calls to UHU_HUH​ in the worst case. We prove a lower bound matching the expected query count; the corresponding worst-case lower bound follows from Somma and de Wolf [SdW26]. The respective bounds on calls to UIU_IUI​ are O(1/γ)O(1/γ)O(1/γ) in expectation and O(γ−1log⁡(1/ε))O(γ^{-1}\log(1/\varepsilon))O(γ−1log(1/ε)) in the worst case. On (N+1)(N+1)(N+1)-dimensional systems, these UIU_IUI​ bounds are also optimal when the expected or worst-case count of UHU_HUH​ calls, respectively, is o((α/Δ)N)o((α/Δ)\sqrt N)o((α/Δ)N​). Both algorithms use a constant-accuracy spectral filter to construct a purifier, which we then sequentially compose during amplitude amplification to prepare a state with constant overlap with the ground state. The expected-query algorithm repeats the preparation followed by one high-accuracy spectral filter until success. The worst-case algorithm uses filters of increasing accuracy and limits the total number of queries.
In the quantum linear systems problem (QLSP), we are given query access to a ddd-sparse N×NN\times NN×N matrix AAA with condition number κκκ, and the ability to prepare a quantum state proportional to a vector b⃗\vec bb. The goal is to output an εεε-approximation to the quantum state proportional to the solution x⃗\vec xx of Ax⃗=b⃗A\vec{x}=\vec{b}Ax=b. Following a long line of work, the best previously known quantum algorithms for the QLSP had query complexities O(κdlog⁡(1/ε))O(κd\log(1/ε))O(κdlog(1/ε)) and κd(κd/ε)o(1)κ\sqrt d(κd/ε)^{o(1)}κd​(κd/ε)o(1), while the best known lower bounds were Ω(κlog⁡(1/ε))Ω(κ\log(1/ε))Ω(κlog(1/ε)) and Ω(κd)Ω(κ\sqrt d)Ω(κd​). We improve these bounds and show that the complexity of the QLSP is Θ(κdlog⁡(1/ε))Θ(κ\sqrt d\log(1/ε))Θ(κd​log(1/ε)). We also resolve an open problem of Berry and Childs by showing that any N×NN\times NN×N unitary can be implemented with bounded error using O(N)O(\sqrt N)O(N​) queries to its matrix entries.