Pasqal 256-Qubit Neutral-Atom Processor Observes Unpredicted Slowdown in 2D Quantum Magnet Dynamics

Pasqal, in collaboration with the Flatiron Institute’s Center for Computational Quantum Physics, the University of Augsburg, Forschungszentrum Jülich, the University of Regensburg, and the Institut d’Optique–CNRS, has mapped the non-equilibrium relaxation landscape of a two-dimensional transverse-field Ising model (TFIM) on a 256-qubit analog neutral-atom processor. Detailed in a preprint titled Exploring the Relaxation Landscape of a 2D Quantum Magnet on a 256-Qubit Processor and an accompanying technical analysis, the study investigates post-quench dynamics following a sudden Hamiltonian parameter shift. The experiments uncover an anomalous relaxation slowdown in a parameter regime where three state-of-the-art classical tensor-network algorithms diverge from one another and fail to maintain controlled late-time predictions.

Rydberg Lattice Quenches and Three Non-Equilibrium Relaxation Regimes The experimental architecture configured 256 neutral rubidium-87 (87Rb) atoms into a 16×16 square array with nearest-neighbor distance R and van der Waals interactions decaying as U(R/rij)6. Ground and Rydberg electronic states encoded the pseudospin degrees of freedom, driven by global two-photon laser excitation with Rabi frequency Ω and detuning δ to simulate ferromagnetic Ising interactions under a transverse field hx. Initializing the processor in a fully polarized product state |↓···↓⟩, researchers executed an abrupt quantum quench across varying ratios of hx/J. The resulting dynamics revealed three distinct physical behaviors: rapid thermalization near the dynamical phase transition (hx/J ≈ 2.5), a long-lived prethermal plateau at large transverse fields (hx/J = 6.0) governed by an emergent U(1)-symmetric XY Floquet Hamiltonian (HZY), and an unexpected relaxation slowdown at intermediate field strengths (hx/J = 3.5). Dynamical Regimes and Classical Simulation Benchmarks on Pasqal 256-Qubit ArrayField Parameter RegimePhysical Behavior Observed on QPUTheoretical & Computational StatusNear-Critical Field (hx/J ≈ 2.5)Rapid thermalization; two-body correlations settle on interaction timescale tJ ≈ 1.Consistent with non-integrable thermalization within a Z2 symmetry-broken sector.Intermediate Field (hx/J ≈ 3.5)Pronounced relaxation slowdown; correlations persist far from thermal values beyond t ≈ 1/J.Unpredicted by existing theory; classical tensor-network methods lose numerical control.Strong Field (hx/J ≈ 6.0)Prethermal plateau; long-lived correlations matching ground state of effective ZY model.Agrees with leading-order Floquet-Magnus expansion in the rotating frame (period TF = π/hx).Classical Simulation Frontier2D Tensor Networks, Matrix Product States, and Tree Tensor Networks diverge at tJ ≳ 1.Entanglement entropy growth saturates classical bond dimensions (χ = 40 to 600).Hardware Noise VerificationSPAM mitigation (ε = 1%, ε’ = 3%), dephasing models, and 5×5 exact diagonalizations.Confirms the three-regime landscape is intrinsic many-body physics, not a machine artifact. Classical Tensor-Network Limits and Physical Validation To verify the experimental observations, the team benchmarked the quantum processor against three numerical methods: two-dimensional tensor networks (2DTN) using belief-propagation message updates, matrix product states (MPS) using two-site time-dependent variational principles (TDVP), and tree tensor networks (TTN). While all three techniques achieved quantitative agreement with hardware measurements during early-time evolution (tJ ≤ 1), they diverged sharply from one another and from the experimental data at later times. Specifically, 2DTN encountered numerical discontinuities from bond-dimension truncations (χ = 40), while MPS (χ = 600) and TTN (χ = 512) introduced ansatz-induced geometric oscillations. In contrast, the neutral-atom processor maintained consistent relaxation trends across scaling lattice dimensions. Noise calibrations incorporating state preparation and measurement (SPAM) corrections and exact state-vector calculations on 5×5 sub-lattices confirmed that the intermediate slowdown originates from intrinsic many-body interactions rather than decoherence. The authors note that the study does not claim a formal computational quantum advantage, as it does not benchmark algorithmic execution runtimes. Instead, the experiment builds upon earlier material validation benchmarks—such as the simulation of the frustrated magnetic crystal TmMgGaO4—to demonstrate that analog neutral-atom arrays can explore non-equilibrium condensed-matter regimes that exceed the convergence boundaries of classical tensor-network simulations. The preprint is indexed on arXiv here, with technical perspectives outlined in Pasqal’s Technical Blog here and official Newsroom Release here. In GQI Portal The players behind the news The team that writes QCR tracks every company, deal and technology in the GQI Factory, GQI's verified database of the quantum industry. Next up: every story linked to its players, coming to QCR's paid plans. Players →Companies and institutions across the quantum industry, by segment. Scorecards →How the players compare on hardware, software, funding and more. Newsletter QCR Alerts in your inbox The latest reporting and analysis from Quantum Computing Report. Free, unsubscribe any time. Sign up for QCR Alerts Leave a commentCancel replyAll fields are required. Your email address will not be published.CommentName Email Type in the text displayed above Δ