Université de Bordeaux Builds Stable Quantum Krylov Method

Researchers at Université de Bordeaux, CNRS, and LOMA have developed a new quantum computing method, Orthogonal Quantum Krylov Diagonalization (OQKD), that directly reformulates the classical Lanczos recursion at the operator level. Unlike existing Quantum Krylov Diagonalization methods, OQKD eliminates the need for overlap-matrix regularization, promising improved numerical stability and accuracy in quantum calculations. The team reports achieving this by expressing Lanczos vectors as polynomial transformations of the Hamiltonian. This increased stability comes without a computational penalty; using block encoding and Generalized Quantum Signal Processing, OQKD achieves the same asymptotic query complexity as established Chebyshev-based QKD methods. Numerical simulations confirm the classical Lanczos convergence and numerical stability of the proposed method. Building upon the OQKD framework, the researchers introduced a restarted state-preparation protocol that replaces a single high-degree polynomial transformation with a sequence of fixed low-degree transformations, maintaining an affordable block encoding success probability while retaining comparable convergence. These results identify the restarted protocol as a promising state-preparation strategy for Quantum Phase Estimation.

A new quantum computing method directly mirrors the stability of a classical algorithm, potentially unlocking more reliable calculations. This replication extends to the convergence properties as well, offering a more predictable and stable pathway to finding low-energy spectra of quantum many-body Hamiltonians. Numerical simulations of the J1-J2 Heisenberg model confirm the classical Lanczos convergence and numerical stability of the proposed method, and the measurement-complexity scaling is established analytically. This suggests potential for further exploration of its application in quantum computing.

Quantum computing’s pursuit of efficient methods for determining the energy levels of complex quantum systems has focused increasingly on subspace-diagonalization techniques, notably Quantum Krylov Diagonalization (QKD). These approaches aim to circumvent the exponential scaling challenges of classical computation by leveraging quantum mechanical principles to generate and manipulate relevant computational spaces. Building upon the OQKD framework, the team introduced a restarted state-preparation protocol that replaces a single high-degree polynomial transformation with a sequence of fixed low-degree transformations, maintaining comparable convergence and an affordable block encoding success probability. These advances establish OQKD as a robust quantum analogue of the classical Lanczos algorithm, offering a promising path toward more reliable and scalable quantum simulations.

Existing approaches, while promising for computing quantum many-body Hamiltonians, necessitate overlap-matrix regularization that limits numerical stability and accuracy. The team aims to resolve this problem with their newly developed Orthogonal Quantum Krylov Diagonalization (OQKD) framework. The core issue is that many QKD methods approximate Krylov vectors without ensuring orthogonality, leading to instability and diminished accuracy. Kirby et al. proposed a quantum Krylov construction based on block encoding and qubitization of Chebyshev polynomials, but the researchers note that “orthogonality between Krylov vectors is still not preserved.” This lack of orthogonality mirrors a challenge faced in classical numerical linear algebra, where ill-conditioning necessitates orthogonalization techniques like the Lanczos algorithm. This maintains comparable convergence and an affordable block encoding success probability.

This means the method requires a comparable number of quantum operations despite its improved handling of numerical errors. Building upon the OQKD framework, the researchers introduced a restarted state-preparation protocol which replaces complex high-degree polynomial transformations with sequences of simpler, fixed-degree transformations, maintaining an affordable block encoding success probability while retaining comparable convergence. These results establish OQKD as an orthogonal quantum analog of the classical Lanczos algorithm and identify the restarted protocol as a promising state-preparation strategy for Quantum Phase Estimation.

This validation confirms the classical Lanczos convergence and numerical stability of the proposed method, while the measurement-complexity scaling is established analytically. Building upon the OQKD framework, the researchers introduced a restarted state-preparation protocol that replaces a single high-degree polynomial transformation with a sequence of fixed low-degree transformations, maintaining comparable convergence and an affordable block encoding success probability. This crucial improvement eliminates the need for overlap-matrix regularization, streamlining calculations and potentially accelerating convergence.

While OQKD directly reformulates the classical Lanczos recursion for improved stability, implementing a single, high-degree polynomial transformation necessary for state preparation can be resource intensive. To circumvent this, the team introduced a restarted state-preparation protocol building upon the OQKD framework, which replaces a single high-degree polynomial transformation with a sequence of fixed low-degree transformations, maintaining comparable convergence and an affordable block encoding success probability. This restarted protocol focuses on sustaining an affordable block encoding success probability, a critical factor in translating theoretical quantum algorithms into practical applications. The approach allows for efficient Krylov state preparation without sacrificing the accuracy gained through OQKD’s orthogonalization techniques, which eliminate the need for overlap-matrix regularization. The development of this protocol is significant because it addresses a key challenge in quantum computation: balancing algorithmic efficiency with the constraints of available quantum resources.

While Quantum Phase Estimation (QPE) aims for precise eigenvalue estimation through controlled time evolution, it’s often hampered by what researchers call the orthogonality catastrophe, where initial state overlap impacts success probability. Variational Quantum Algorithms (VQAs), designed for near-term processors, struggle with optimization challenges like barren plateaus, noise sensitivity, and dependence on the expressibility of the chosen ansatz. OQKD, however, takes a different approach, focusing on constructing orthogonal Krylov subspaces to improve numerical stability. This eliminates the need for overlap-matrix regularization, a significant step towards more reliable quantum calculations. Building on the OQKD framework, the researchers introduced a restarted state-preparation protocol, replacing complex polynomial transformations with sequences of lower-degree ones. This maintains an affordable block encoding success probability while retaining comparable convergence. These results establish OQKD as an orthogonal quantum analog of the classical Lanczos algorithm and identify the restarted protocol as a promising state-preparation strategy for Quantum Phase Estimation.