A new algorithm called Sweeping RTM enables approximate calculation of output probabilities from one-dimensional chaotic quantum circuits at a specified level of precision. The method utilises transverse tensor networks based on reduced transition matrices to compress information efficiently during computation; Matilde Grassi, Stefano Carignano, Luca Tagliacozzo, and Jacopo De Nardis led the research. A new computing method simulates complex quantum systems using standard computers effectively.

The team focused on calculating how likely specific outcomes are in these simulations, streamlining computation via compression techniques utilising reduced transition matrices to manage information efficiently. These advancements allow classical simulation of certain chaotic quantum circuits previously considered impossible due to increasing computational demands. Researchers 8089, CY Cergy Paris Université, alongside colleagues at Affiliation: Barcelona Supercomputing Centre, devised a new method for simulating complex quantum systems using conventional computers.

Their approach focuses on calculating the probability of specific outcomes within these simulations with defined accuracy. Colleagues also work at Affiliation: Institute of Fundamental Physics IFF-CSIC and Quantum Advanced Research Centre (QuARC), PSL Research University.

The team tackled a longstanding problem in quantum computing where entanglement rapidly increases computational demands as system complexity grows. To address this, they utilise ‘tensor networks’, representing complicated quantum states as interconnected simpler components. This allows them to compress information efficiently during calculations and focus on evaluating probabilities rather than tracking every step of evolution, specifically employing reduced transition matrices to manage data effectively.

Reduced computational cost enables accurate simulations of extended quantum chaos

Scientists, collaborating with researchers from Barcelona Supercomputing Centre and Institute of Fundamental Physics IFF-CSIC, have achieved subexponential scaling of bond dimension, a measure of computational effort, when estimating output probabilities. This improvement addresses limitations in prior tensor network methods which were hampered by extensive spatial entanglement growth. Previously, an exponential increase in necessary resources limited the accuracy when simulating chaotic one-dimensional brick-wall circuits as simulation time grew; now, stable probability estimates are attainable for complex systems thanks to this reduced demand on computing power.

Retained singular values, representing key information during compression, exhibit bounded or logarithmic growth according to the team. Exponential increases observed with full quantum state mapping contrast sharply with these findings. Analysis of sixty-qubit circuits revealed that output-probability statistics converged towards predictions established by Porter and Thomas for log moments, validating algorithmic accuracy through indirect means. However, current observations provide heuristic evidence only and do not definitively establish an asymptotic complexity limit, practical application to even larger systems therefore remains challenging.

Demonstrating probabilistic simulation benefits from limited dimensional complexity and numerical precision

Subexponential scaling in estimating output probabilities offers a potential route toward classically simulating chaotic quantum systems; however, the present method is constrained to one-dimensional brick-wall circuits using fixed precision levels. Whether this observed efficiency will translate to more complex circuit architectures or if demanding higher accuracy would negate these gains requires further investigation. It’s important to acknowledge that demonstrating subexponential scaling occurred on simplified one-dimensional circuits with a specific level of accuracy, as extending these results to realistic three-dimensional systems, or requiring greater precision, could prove challenging.

Subexponential growth in required computational effort represents an advance over previous methods hampered by rapidly increasing entanglement during simulation time. The Sweeping Reduced Transition Matrix algorithm offers an innovative approach for simulating chaotic quantum systems; it focuses on calculating output probabilities instead of tracking complete state evolution and streamlines calculations for complex simulations. This new method provides benefits beyond simply reducing the resources needed, potentially opening doors to exploring previously inaccessible regimes within quantum chaos research.

Estimating output probabilities from one-dimensional brick-wall circuits demonstrated subexponential scaling in computational effort with this new technique. This means that classically simulating these specific chaotic quantum systems may require fewer computing resources than previously thought because the complexity does not increase as rapidly over time. Researchers used a transverse tensor network contraction called the Sweeping Reduced Transition Matrix algorithm to achieve this result with up to sixty qubits at fixed precision levels. The authors note further work is necessary to determine if these findings extend to more complex circuit designs or higher accuracy requirements.

👉 More information
🗞 Strong Simulation of 1D Quantum Circuits via Reduced Transition Matrices
✍️ Matilde Grassi and Jacopo De Nardis (CY Cergy Paris Université); Stefano Carignano (Affiliation: Barcelona Supercomputing Center); Luca Tagliacozzo (Affiliation: Institute of Fundamental Physics IFF-CSIC)
🧠 ArXiv: https://arxiv.org/abs/2610.02082

More like this