Xinzhao Wang of Peking University and colleagues have established a new scaling relationship for simulating open quantum systems. Their work derives commutator-based error bounds for Lindbladian simulation, achieving a square root of N (√N) scaling in the number of Trotter steps for systems with N sites. This represents an improvement over previous simulation techniques, particularly for locally interacting systems. The researchers also applied Richardson extrapolation, reducing the number of steps needed per circuit to depend polylogarithmically on the desired precision while maintaining the improved scaling.
Lindbladian Simulation via Trotter Decomposition
This advance centers on refining the Trotter decomposition method, a standard approach to approximating how quantum systems evolve over time when interacting with their environment. The research specifically addresses a longstanding challenge in accurately bounding the errors that accumulate during these simulations. These bounds allow for a more precise prediction of simulation accuracy, enabling researchers to reduce the computational resources needed to achieve a desired level of precision.
To further enhance efficiency, the researchers applied Richardson extrapolation, a technique borrowed from numerical analysis, to estimate observable averages. The work demonstrates that the number of steps needed scales favorably with increasing precision, a critical factor for tackling larger, more realistic simulations. A key component of this work is a new truncation bound for the Baker-Campbell-Hausdorff (BCH) expansion, a mathematical tool used to approximate the composition of quantum operations.
Existing BCH expansions often suffer from convergence issues in physically relevant systems, limiting their applicability. The researchers’ new bound bypasses these issues, providing a more robust and reliable way to control the error introduced by the approximation. The study’s findings demonstrate that Trotter-based methods outperform prior simulation techniques in terms of system-size scaling while maintaining a low overhead in terms of auxiliary quantum bits, specifically, requiring only O(1) ancillas.
Numerical simulations, conducted to validate the theoretical results, support the predicted scaling behavior, confirming the efficiency gains. The researchers tested their approach on Ising-chain and Heisenberg-chain Lindbladian simulations with various initial states and dissipation strengths. Numerical simulations further support the predicted system-size and precision scaling. For example, simulations with initial state and local relaxation demonstrate the effectiveness of the approach across different system sizes. The team also examined the impact of local amplitude damping and phase damping, further validating the robustness of the method.
Commutator Bounds for Lindbladian Error Scaling
Establishing tighter bounds on the errors inherent in simulating open quantum systems allows for more efficient calculations of system behavior as complexity increases. The work details a method for estimating these errors based on nested commutators, yielding a scaling advantage for locally interacting systems with N sites, specifically, a reduction in the number of computational steps needed to achieve a given level of accuracy. This contrasts with prior approaches that relied on summing the norms of individual operator components, a method less sensitive to the underlying structure of quantum dynamics.
The new approach centers on a commutator-based error bound, offering a complexity scaling governed by nested commutators rather than the sum of operator norms. This is particularly significant because it addresses a gap in existing literature, where such bounds were largely absent for Lindbladian simulations, despite being well-established for Hamiltonian systems.
Baker-Campbell-Hausdorff Truncation for Remainder Bounds
The research, detailed in recent work, focuses on refining the precision of Trotter-based methods, a common approach to approximating the complex evolution of quantum states. The core of this improvement lies in a novel approach to managing the accumulation of errors inherent in the Trotter decomposition.
By carefully bounding the remainder after truncating the Baker-Campbell-Hausdorff expansion, Xinzhao Wang, Shuo Zhou, Xiaoyang Wang, Yi-Cong Zheng, Shengyu Zhang, and Tongyang Li have created a method that avoids the divergence issues that often plague high-precision simulations. This is particularly important for Lindbladian dynamics, which describes the evolution of open quantum systems interacting with their environment, where existing error bounds derived from Hamiltonian simulation did not readily translate.
The team’s analysis also addresses the challenges associated with nested commutators, which govern the accuracy of Trotter-based simulations. By carefully controlling the growth of these commutators, they have derived bounds on the simulation error that scale favorably with the system size. The analysis considers the impact of local interactions, represented by the parameter ‘k’ and the extensiveness of the Lindbladian, denoted by ‘g’, to provide a comprehensive understanding of the simulation’s performance.
The work highlights that the factorial growth within inner commutators is controlled by bounding the grade of the expansion, ensuring that the error remains scalable with system size. The development of a general truncation bound for the BCH expansion not only improves the accuracy of simulations but also provides a valuable tool for analyzing the convergence of high-precision Trotter methods, benefiting both open and closed quantum systems.
Locality and the (Γ,k)-Local Lindbladian Definition
The analysis of (Γ,k)-local Lindbladians establishes a relationship between the grade of nested commutators and the accuracy of open quantum system simulations, a connection previously poorly understood. Specifically, the work demonstrates that bounding the growth of these commutators yields a simulation error scaling with the square root of the number of sites, denoted as √N, for locally interacting systems.
The work further specifies that for a (Γ,k)-local and g-extensive Lindbladian on an N-site lattice, the number of Trotter steps needed scales as O( √(N)(kgt)^(3/2)ε^(-1/2)), where ε represents the desired precision and g is the extensiveness parameter. Extrapolation techniques are then applied to further refine the accuracy of observable averages, reducing the dependence of Trotter steps per circuit to a polylogarithmic relationship with the inverse precision.
The extrapolation algorithm, as detailed in the paper, estimates tr[] to precision ε|O| with a probability of at least 2/3, using O(1/ε^2) circuit runs, each consisting of O( √(N)(kgt)^(3/2)) Trotter steps. The notation O(·) hides polylogarithmic factors, indicating a potentially substantial reduction in computational demands for high-precision simulations. This bound bypasses common convergence issues encountered in physically relevant systems. The summation over the total grade, combined with Theorem 2, provides a detailed framework for controlling this error.
The implementation of these local channels within each Trotter step can be achieved using Stinespring dilation, requiring a unitary acting on k system qubits and at most 2k ancillas, followed by tracing out the ancillas. These ancillas can be reused, minimizing the overhead associated with maintaining a large number of qubits.
The analysis considers systems with block encodings of Hamiltonian terms and local jump components, with normalization factors = Θ(||) and = Θ(||), respectively. The LCU construction provides a block encoding of = using O(Γ) queries and gates.
👉 More information
🗞 Lindbladian simulation with commutator bounds
✍️ Xinzhao Wang, Shuo Zhou, Xiaoyang Wang, Yi-Cong Zheng, Shengyu Zhang and Tongyang Li
🧠 DOI: http://link.aps.org/doi/10.1103/yt2x-h33t




See today’s quantum computing news on Quantum Zeitgeist for the latest breakthroughs in qubits, hardware, algorithms, and industry deals.
