Trotterisation Thresholds and Dissipative Quantum Chaos in Open-System Digital Quantum Simulation

AI summary

Overview

Research area: Quantum information science — digital quantum simulation (DQS), open quantum systems, and quantum chaos.

Technical level: Advanced. The paper assumes familiarity with Lindblad master equations, Trotter–Suzuki decompositions, Kraus/operator-sum representations, and random matrix theory (RMT).

Scope (one sentence): The paper asks whether the "Trotterisation threshold" — the step size beyond which gate-based quantum simulation abruptly fails because of quantum chaos — survives when the simulated system is open, dissipative and decohered, and it builds RMT-based statistical tools to characterise the resulting chaos.

What This Paper Is About

Digital quantum simulation approximates continuous quantum dynamics with a sequence of discrete gates, and the size of each Trotter step trades gate count against accuracy. In closed systems it is known that increasing the step size past a threshold causes a sudden collapse in accuracy traced to the onset of quantum chaos, but this had only been studied without environmental noise and decoherence — exactly the conditions real quantum computers never enjoy. The authors extend that analysis to open-system simulators, asking whether the threshold survives dissipation, whether it is visible in the long-time steady state after coherence is lost, and how to rigorously diagnose quantum chaos when the evolution operator is no longer unitary.

Key Contributions

1. Demonstration that the Trotterisation threshold persists in open-system DQS. Across three paradigmatic models with physically motivated dissipation channels, the threshold remains at essentially the same Trotter step size as in the closed-system case over a broad range of dissipation strengths, vanishing only when dissipation is strong enough to dominate the dynamics even on the timescale of a single Trotter step.

2. Observation of the threshold in the steady-state limit. The threshold remains detectable in steady-state expectation values of local observables, showing that surviving coherence is not required to see it — visibility depends on the balance between unitary and dissipative dynamics.

3. New RMT-based statistical diagnostics for open systems. The authors introduce a "residual unitary quantum chaos" analysis based on the dominant (canonical) Kraus operator $K_1$ of the Floquet superoperator, together with criteria for handling inherited symmetries: a rotated ($\sigma_x$) basis for the Heisenberg model to erase residual parity structure, and exact parity-block bookkeeping for the Rabi model.

4. Identification of the conditions for genuinely dissipative quantum chaos. Dissipative chaos emerges only when dissipation is strong enough to induce genuinely dissipative behaviour but not so strong as to suppress the underlying unitary chaotic structure — the threshold for it sits at the same Trotter step size as the unitary one.

Main Findings

• Three dynamical regimes survive dissipation: at small Trotter step sizes the observables show periodic or quasi-periodic revivals that decay faster with stronger dissipation; at intermediate step sizes the revival period shifts and revivals weaken; beyond the threshold quasi-periodicity is lost entirely and the time evolution becomes irregular, accompanied by a sharp drop in simulation fidelity. The transition is sharp enough to identify by visual inspection.

• Threshold position is set by the unitary dynamics: comparing simulation runs at dissipation scaling factors $f_\gamma = 0.01, 1, 10, 30, 100$ (applied as $\gamma_k \to f_\gamma \gamma_k$) shows no obvious shift in threshold position until dissipation is strong. The threshold becomes less pronounced, and eventually vanishes, as dissipation grows — earlier for the Rabi model because of the amplified impact of dissipation rates in the cavity.

• Steady-state signatures: for all three models the steady-state observable values show a sharp transition versus Trotter step size, provided $f_\gamma$ is not too large. Beyond a critical dissipation strength the threshold shifts to larger step sizes before disappearing; in the post-threshold regime, larger step sizes require smaller dissipation for the post-threshold patterns to vanish. Stable-island behaviour (vertical stripes in local observables and fidelity) appears post-threshold, as previously identified in the closed-system work.

• The Heisenberg model behaves differently at very large dissipation: its threshold persists even at very large dissipation rates, though shifted toward smaller Trotter step sizes. This is attributed to $\pi/2$ rotation gates in the Trotterised evolution reorienting the dissipation operators, leaving a non-trivial steady state determined by competition between dissipation channels acting along different axes — unlike the kicked top and Rabi models, which are driven toward a simple ground state.

• Open DQS can faithfully reproduce open target dynamics: at $f_\gamma = 1$, fidelities against an effective Lindblad open target remain high for long times across all models, despite there being no guarantee that the Floquet superoperator admits an effective Lindbladian. Fidelity against a closed-system target decays rapidly in the presence of dissipation, as expected.

• Symmetry inheritance in the dominant Kraus operator: at moderate dissipation in the post-threshold regime — $\tau = 0.5$, $f_\gamma = 1$ (kicked top); $\tau = 0.2$, $f_\gamma = 1$ (Heisenberg); $\tau = 0.1$, $f_\gamma = 0.1$ (Rabi) — the $K_1$ eigenvector statistics of all three models most closely match COE distributions, consistent with the time-reversal symmetry of their unitary limits.

• Approximate unitarity of $K_1$: the Kraus weight $d_1/D$ approaches unity and the unitarity measure $|K_1^\dagger K_1 - I|/D$ approaches zero when dissipation is weak and the Trotter step is small, justifying the residual-unitary-chaos diagnostic in that regime. The analysis of the onset of this behaviour, and of genuinely dissipative chaos, is where the provided text ends; the paper's later sections are not included in the supplied content.

Methodology in Plain English

The authors model environmental noise with the Markovian Lindblad master equation, assigning each Trotter gate its own open-system propagator, and multiply these into a Floquet superoperator. They deliberately reuse the three models and Trotterisation protocols from their earlier closed-system study — the quantum kicked top (spin $j=10$), a nearest-neighbour Heisenberg spin chain ($N=5$ qubits) and the Rabi model (cavity truncation $\dim_c = 20$) — and add physically motivated dissipation channels with reference rates chosen for relevance to current circuit QED implementations, then scale all rates by $f_\gamma$ to sweep dissipation strength.

For quantum chaos diagnostics, they borrow the logic of random matrix theory: chaotic quantum dynamics have universal eigenvalue and eigenvector statistics fixed by symmetry class rather than microscopic detail. In a closed system the evolution is unitary and its eigenvalues lie on the unit circle. In an open system the evolution is a superoperator with eigenvalues inside the unit circle, so the closed-system tools cannot be applied directly. The authors' workaround is to extract the dominant Kraus operator $K_1$ from the operator-sum representation of the Floquet map — the single branch carrying most of the probability — and apply their established eigenvector-statistics test to it, measured by a goodness-of-fit statistic $X^2_{\mathrm{RMT}}$ judged against a 99% confidence region of the appropriate reduced chi-squared distribution. Where inherited parity symmetry distorts the statistics, they either rotate into an incompatible basis (Heisenberg) or analyse only the non-zero parity blocks with $D \to D/2$ and cavity truncation $\dim_c = 4$ (Rabi).

Why This Matters

The result that the threshold is robust to noise, and observable in steady-state values of local observables, means experimentalists can locate it without needing long coherence times or full tomography — relevant because real hardware is always noisy and because smaller Trotter steps cost more gates, so knowing the true usable step size directly sets resource requirements. The work also reframes dissipation: rather than only a nuisance, it can be the target of the simulation, and the fidelity results suggest digitised platforms can reproduce non-digitised open-system dynamics.

Applications implied by the paper:

• Selecting Trotter step sizes and gate budgets for simulation on noisy hardware, using the threshold as a hard accuracy boundary.
• Simulating open-system target models, including effective Lindblad dynamics, on a digital simulator subject to its own decoherence.
• Engineered dissipation for preparing and stabilising entangled states, squeezed states and many-body steady states.
• Using steady-state local observables as a practical, low-cost experimental diagnostic of chaotic breakdown.

Industry relevance: the reference dissipation rates are explicitly chosen for relevance to current circuit QED implementations, and the co-author affiliations include a quantum computing company, indicating direct relevance to superconducting-qubit hardware vendors and cloud quantum computing providers.

Future Directions

• Experimental confirmation on hardware: whether the steady-state threshold signature can be located on a real circuit QED device at the reference dissipation rates used here.
• Extending the analysis beyond the weak-coupling, Markovian regime where the Lindblad description holds, and testing whether the qualitative picture survives non-Markovian noise.
• Generalising the residual-unitary and dissipative-chaos diagnostics to other symmetry classes, other dissipation channels and larger systems than $j=10$, $N=5$ and $\dim_c=20$.
• Determining whether engineered dissipation can be designed so that dissipative quantum chaos itself becomes a resource — for example for state stabilisation — rather than a failure mode.

Target Audience

Quantum information researchers working on digital quantum simulation and Trotterisation error; open quantum systems theorists interested in chaos and random matrix theory beyond Hermitian settings; and experimentalists on superconducting or trapped-ion platforms who need to choose Trotter step sizes and understand where noise-driven accuracy breakdowns occur. The paper is not suitable as an introductory read — the RMT and Kraus-operator machinery presumes graduate-level quantum physics.

Authors’ abstract

Digital quantum simulation (DQS) is one of the most promising applications of quantum computing. Trotterised DQS models quantum dynamics by decomposing the evolution into a sequence of gate-based Trotter steps. Using larger Trotter steps reduces the required gate count, but it is now understood that the Trotterisation step size cannot be increased beyond a certain threshold value, beyond which there is a sudden breakdown in simulation accuracy, due to the onset of quantum chaos. To date, this behaviour has only been studied for closed systems not subjected to the environmental noise and decoherence to which real-world quantum computers are highly sensitive. In this work, we study a range of experimentally relevant open-system quantum simulator models and show that Trotterisation thresholds persist even when open-system effects are included, even in the long-time, steady-state limit, once all coherence has been lost. We show that open DQS can accurately reproduce the dynamics of open target systems prior to the threshold, but still exhibits a sudden performance breakdown beyond it. We introduce new, statistical techniques to rigorously characterise the onset of quantum chaos beyond the Trotterisation threshold for both approximately unitary and genuine dissipative quantum chaos, based on random matrix theory (RMT), building on techniques developed for closed systems. Finally, we demonstrate that dissipative quantum chaos emerges when the dissipation is in balance with the presence of underlying unitary quantum chaos; the dissipation must be strong enough to induce genuinely dissipative behaviour, yet not so strong as to suppress the underlying chaotic structure.