| name | quantum-linear-solver-beyond-condition |
| version | 1.0.0 |
| description | Quantum linear system solving methodology that overcomes the condition number barrier. Uses truncation-based and filtering-based solvers with complexity independent of worst-case condition number kappa. Introduces effective condition number bounds and affine dilation input model. |
| category | quantum |
| tags | ["quantum","linear-systems","algorithms","numerical-analysis","condition-number","quantum-algorithms"] |
| trigger_words | ["quantum linear system solver","quantum HHL algorithm","quantum linear equations","condition number quantum","quantum algorithm linear system","block encoding linear system","effective condition number"] |
| source_paper | arXiv:2607.07691 - Faster quantum linear system solver beyond the condition number (2026) |
Quantum Linear System Solver Beyond the Condition Number
Overview
Two quantum algorithms for solving linear systems Ax = b with query complexity independent of the worst-case spectral condition number κ = ||A^{-1}||. Both solvers produce the normalized quantum state |x⟩ to accuracy ε, dramatically improving upon the standard O(κ) dependence.
Core Methodology
Input Models
- Standard Block Encoding Model:
A accessed via block encoding, |b⟩ prepared by unitary
- Affine Dilation Model (novel): Encodes
A and |b⟩ jointly, enabling further query complexity refinements
Solver 1: Truncation-Based
Query complexity to A:
O(κ_eff · polylog(κ_eff / ε))
Query complexity to |b⟩: Optimal
Effective Condition Number Bounds
For positive even integer t:
κ_eff ≤ ||(A†A)^{-t/2} |x⟩||^{1/t} / ε^{1/t}
For positive odd integer t:
κ_eff ≤ ||A^{-1†} (A†A)^{-(t-1)/2} |x⟩||^{1/t} / ε^{1/t}
Solver 2: Filtering-Based
When solution norm is known:
Query complexity = 6 · ||A^{-1†} |x⟩|| / ε · ln(1/ε)
Extremely simple implementation with favorable runtime prefactor.
Key Innovation: The κ-Barrier Breakthrough
Traditional quantum linear system solvers (e.g., HHL) have complexity scaling as O(κ). This is a worst-case measure that can significantly overestimate runtime for typical instances. The new solvers achieve complexity dependent on κ_eff, which can be much smaller than κ for well-conditioned solution states.
Solution Norm Estimator
A similarly simple estimator with the same asymptotic cost (up to logarithmic factors) for cases where ||x|| is unknown.
Activation
This skill activates when designing quantum linear system algorithms, analyzing quantum algorithm complexity, implementing HHL-type solvers, or studying quantum numerical linear algebra.
Related Concepts
- Block encoding techniques
- Quantum singular value transformation (QSVT)
- Effective vs spectral condition numbers
- Quantum state preparation
- Quantum algorithm complexity analysis