We give the first sublinear-time classical solvers for sparse semidefinite programs in the bounded-radius regime, without low-rank assumptions or Frobenius norm dependence on the constraint matrices. For constant precision and bounded primal and dual radii, prior quantum algorithms of Brandão et al. (2019) and van Apeldoorn and Gilyén (2019) achieved $\widetilde{O}(\sqrt{n}+\sqrt{m})$ dependence on matrix dimension $n$ and constraint number $m$. Compared with the $\widetilde{O}(mn)$ runtime of existing classical methods, this suggests a quartic quantum speedup when $m \approx n$. Beyond a usual Grover speedup, this separation relies on the Quantum OR lemma, whose sample-reuse mechanism decouples the cost of Gibbs-state preparation from constraint search. We show that this reuse mechanism is classically realizable for sparse SDPs.
Our main technical contribution is a classical procedure for simultaneously estimating many expectation values with respect to a sparse Hamiltonian's Gibbs state. This combines randomized Lánczos filtering with an efficient sampling-based estimator. We also introduce a stochastic online-learning framework for SDP solving, substantially improving accuracy-dependence over standard oracle-based MMWU approaches. Let $s$ denote the the input matrix sparsity and $γ:=Rr/\varepsilon$ capture dependence on the primal $(R)$ and dual $(r)$ radii as well as target accuracy $(\varepsilon)$. When $γ^2\leq\min\{m,n/s\}$, our solver runs in time $\widetilde{O}\left(nsγ^{4.5}+msγ^2\right)$. For $γ=O(1)$, this is $\widetilde{O}\left((n+m)s\right)$ and sublinear in the $O(mns)$ input size. Similar to the quantum algorithms, this matches known lower bounds with respect to $m$ and $n$, up to logarithmic factors. This implies that, with respect to dimensions $m$ and $n$, there is no super-quadratic quantum advantage for generic sparse SDP solving.
We give the first sublinear-time classical solvers for sparse semidefinite programs in the bounded-radius regime, without low-rank assumptions or Frobenius norm dependence on the constraint matrices. For constant precision and bounded primal and dual radii, prior quantum algorithms of Brandão et al. (2019) and van Apeldoorn and Gilyén (2019) achieved O(n+m) dependence on matrix dimension n and constraint number m. Compared with the O(mn) runtime of existing classical methods, this suggests a quartic quantum speedup when m≈n. Beyond a usual Grover speedup, this separation relies on the Quantum OR lemma, whose sample-reuse mechanism decouples the cost of Gibbs-state preparation from constraint search. We show that this reuse mechanism is classically realizable for sparse SDPs.
Our main technical contribution is a classical procedure for simultaneously estimating many expectation values with respect to a sparse Hamiltonian's Gibbs state. This combines randomized Lánczos filtering with an efficient sampling-based estimator. We also introduce a stochastic online-learning framework for SDP solving, substantially improving accuracy-dependence over standard oracle-based MMWU approaches. Let s denote the the input matrix sparsity and γ:=Rr/ε capture dependence on the primal (R) and dual (r) radii as well as target accuracy (ε). When γ2≤min{m,n/s}, our solver runs in time O(nsγ4.5+msγ2). For γ=O(1), this is O((n+m)s) and sublinear in the O(mns) input size. Similar to the quantum algorithms, this matches known lower bounds with respect to m and n, up to logarithmic factors. This implies that, with respect to dimensions m and n, there is no super-quadratic quantum advantage for generic sparse SDP solving.
We study quantum algorithms for approximating Lasserre's hierarchy values for polynomial optimization. Let f,g1,…,gm be real polynomials in n variables and f⋆ the infimum of f over the semialgebraic set S(g)={x:gi(x)≥0}. Let λk be the value of the order-k Lasserre relaxation. Assume either (i) f⋆=λk and the optimum is attained in the ℓ1-ball of radius 1/2, or (ii) S(g) lies in the simplex {x≥0:∑jxj≤1/2}, and the constraints define this simplex. After an appropriate coefficient rescaling, we give a quantum algorithm based on matrix multiplicative weights that approximates λk to accuracy ε>0 with runtime, for fixed k, O(nkε−4+nk/2ε−5),O(sg[nkε−4+(nk+∑i=1mnk−di)1/2ε−5]), where sg bounds the sparsity of the coefficient-matching matrices associated with the constraints. Classical matrix multiplicative-weights methods scale as O(n3kpoly(1/ε)) even in the unconstrained case. As an example, we obtain an O(nε−4+nε−5) quantum algorithm for portfolio optimization, improving over the classical O(nω+1log(1/ε)) bound with ω≈2.373.
Our approach builds on and sharpens the analysis of Apeldoorn and Gilyén for the SDPs arising in polynomial optimization. We also show how to implement the required block encodings without QRAM. Under the stated assumptions, our method achieves a super-quadratic speedup in the problem dimension for computing Lasserre relaxations.
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 ε. The dependence on κ and ε is already well understood through the lower bound Ω(κlog(1/ε)), 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(κslog(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) for constant error ε. In this work, we complete the picture and prove the full joint lower bound Ω(κslog(1/ε)) in the sparse-access model, thereby establishing the anticipated dependence on all three parameters simultaneously.
We give quantum algorithms for Hamiltonian simulation, linear differential equations, and linear systems with square-root dependence on sparsity. For an s-sparse Hermitian matrix with row and column Euclidean norms bounded by η, evolution for time t to error ε uses O(sτ+sτlog(1/ε)+log(1/ε)) queries, where τ=η∣t∣. The same bound holds for time-dependent Hamiltonians with coherent time-indexed sparse access and certified time discretisation. Our construction combines an exact entrywise representation with Cayley transducers. A joint lower bound nearly matches the time, sparsity, and precision dependence, ruling out a uniformly additive bound in sτ and log(1/ε). For dissipative linear differential equations we obtain O(gˉ[s(τ+τ)+1]log(Cgˉ/ε)) queries, where τ is the evolution time multiplied by a norm bound, gˉ bounds the ratio of combined input and source norms to the final-solution norm, and C is an absolute constant. We sharpen this bound by separating Hamiltonian and dissipative components, and treat general constant generators. Lower bounds capture joint parameter dependence for time-dependent equations and precision hardness for constant positive semidefinite dissipation. For linear systems, a Cayley walk and Chebyshev filter give O(κslog(1/ε)) queries when ∥A∥≤1 and ∥A−1∥≤κ, and we give an alternative proof of a matching joint lower bound. We also give gate-efficient implementations of time-independent Hamiltonian simulation and the dissipative ODE solver, including Lipschitz time dependence, retaining the query bounds with polylogarithmic overhead in inverse precision given efficient entry decoding, arithmetic, and required source preparations.
Quadratic Unconstrained Binary Optimization (QUBO) problems are prevalent in various applications and are known to be NP-hard. The seminal work of Goemans and Williamson introduced a semidefinite programming (SDP) relaxation for such problems, solvable in polynomial time that upper bounds the optimal value. Their approach also enables randomized rounding techniques to obtain feasible solutions with provable performance guarantees. In this work, we identify instances of QUBO problems where matrix multiplicative weight methods lead to quantum and quantum-inspired algorithms that approximate the Goemans-Williamson SDP exponentially faster than existing methods, achieving polylogarithmic time complexity relative to the problem dimension. This speedup is attainable under the assumption that the QUBO cost matrix is sparse when expressed as a linear combination of Pauli strings satisfying certain algebraic constraints, and leverages efficient quantum and classical simulation results for quantum Gibbs states. We demonstrate how to verify these conditions efficiently given the decomposition. Additionally, we explore heuristic methods for randomized rounding procedures and extract the energy of a feasible point of the QUBO in polylogarithmic time. While the practical relevance of instances where our methods excel remains to be fully established, we propose heuristic algorithms with broader applicability and identify Kronecker graphs as a promising class for applying our techniques. We conduct numerical experiments to benchmark our methods. Notably, by utilizing tensor network methods, we solve an SDP with D=250 variables and extract a feasible point which is certifiably within 0.15% of the optimum of the QUBO through our approach on a desktop, reaching dimensions millions of times larger than those handled by existing SDP or QUBO solvers, whether heuristic or rigorous.
We give a quantum algorithm for simulating a d-sparse Hermitian Hamiltonian H, assuming a known upper bound Λ on its maximum column Euclidean norm ∥H∥1→2. For tΛ≥1/2, simulation with operator-norm error ε uses O(tΛd+dlog(2/ε)) sparse-oracle queries. This removes the subpolynomial overhead in Low's algorithm [STOC 2019], replacing it with an additive logarithmic precision term. For d>1 and tΛ≥log(2/ε), the bound matches the worst-case lower bound. A known spectral-norm upper bound may also be used in place of Λ. The number of 1- and 2-qubit gates is linear in the query scale, up to oracle costs and polynomial overhead in the input bit lengths and logarithmic precision parameters.
As applications, we obtain O(κdpolylog(κ/ε)) queries for solving d-sparse quantum linear systems with ∥A∥≤1 and ∥A−1∥≤κ, under standard sparse and state-preparation access. We also give a gate-efficient implementation of black-box unitaries with at most d nonzero entries per row and column using O(dlog(2/ε)) queries, given sparse access to the unitary and its adjoint. At constant error, the query bound is optimal and yields Θ(N) queries for arbitrary N×N unitaries, resolving the open question on black-box unitary implementation posed by Berry and Childs [QIC 2012].
Solving optimization problems is a key task for which quantum computers could possibly provide a speedup over the best known classical algorithms. Particular classes of optimization problems including semidefinite programming (SDP) and linear programming (LP) have wide applicability in many domains of computer science, engineering, mathematics, and physics. Here we focus on semidefinite and linear programs for which the dimensions of the variables involved are exponentially large, so that standard classical SDP and LP solvers are not helpful for such large-scale problems. We propose the QSlack and CSlack methods for estimating their optimal values, respectively, which work by (1) introducing slack variables to transform inequality constraints to equality constraints, (2) transforming a constrained optimization to an unconstrained one via the penalty method, and (3) replacing the optimizations over all possible nonnegative variables by optimizations over parameterized quantum states and parameterized probability distributions. Under the assumption that the SDP and LP inputs are efficiently measurable observables, it follows that all terms in the resulting objective functions are efficiently estimable by either a quantum computer in the SDP case or a quantum or probabilistic computer in the LP case. Furthermore, by making use of SDP and LP duality theory, we prove that these methods provide a theoretical guarantee that, if one could find global optima of the objective functions, then the resulting values sandwich the true optimal values from both above and below. Finally, we showcase the QSlack and CSlack methods on a variety of example optimization problems and discuss details of our implementation, as well as the resulting performance. We find that our implementations of both the primal and dual for these problems approach the ground truth, typically achieving errors of the order 10-2.