Photonic quantum computing models fluid flows

The team employed a technique called bright scattershot boson sampling, extending previous work to use a wider range of photon numbers and overcome limitations in scaling. This reconfigurable architecture accurately reproduces the temporal dynamics of systems ranging from two-dimensional convection to global atmospheric circulation, remaining resilient to photon losses of up to 50 percent. The work demonstrates that this photonic approach exceeds the prediction horizon of a classical reservoir-computing model with comparable computational demands.

Bright Scattershot Boson Sampling for Fluid Dynamics

This modeling capability stems from a reconfigurable architecture programmed to autonomously simulate the time evolution of nonlinear dynamical systems. The team extended a technique called scattershot boson sampling by incorporating higher photon number components, a new scheme they term bright scattershot boson sampling, or BSBS, to address scaling limitations inherent in the original approach.

BSBS uses all randomly generated input photon number distributions, a key distinction from earlier methods, and allows for the exploitation of high-dimensional correlations generated by multiphoton measurements. The researchers verified the superiority of BSBS by comparing its performance to a run using coherent light, revealing a clear advantage for the quantum light source, particularly in the higher-dimensional model simulating atmospheric circulation.

High-Dimensional Correlations in Multiphoton Boson Sampling

The ability to model complex fluid dynamics using a photonic quantum computer stems from exploiting the high-dimensional correlations inherent in multiphoton boson sampling, a technique extended by researchers to overcome limitations of earlier approaches. This advancement moves beyond simply performing calculations, demonstrating a capacity to simulate physical systems with a reconfigurable architecture programmed to autonomously model time-dependent behavior.

A digital twin of a small-scale multiphoton boson sampler, simulated on a high-performance computer, allowed for the prediction of two nonlinear dynamical systems, one with eight degrees of freedom modeling two-dimensional thermally driven convection and another with 16 dimensions representing global atmospheric circulation.

This simulation-based approach enabled a direct comparison with a classical reservoir computing model, revealing a significant performance advantage for the photonic system. Specifically, the photonic reservoir computing model achieved a prediction horizon twice as long as its classical counterpart for the more complex learning task, outlining a pathway toward practical dynamical prediction using scattershot boson sampling.

Scaling BSBS with Higher Photon Number Components

A digital twin of the BSBS architecture, built on multiphoton statistics, facilitated a direct comparison with a classical analog using coherent light as an input. Simulations revealed that the BSBS-based reservoir computing model outperformed its coherent light counterpart, suggesting a tangible benefit from harnessing the intrinsically high-dimensional correlations generated by the multiphoton boson sampling machine. The paper states that “The BSBS approach has been designed, such that processing of the samples is straightforward,” highlighting the efficiency of the method.

The simulations account for Gaussian states, enabling efficient modeling of state generation, interferometer action and photon loss, while calculating Hafnians, a computationally intensive process, with dimensions linear to the order of photon number moments. While the current digital twin simulations do not yet fully capture the statistical noise introduced by transitioning from calculated expectation values to sampled photon number moments, the design of the BSBS approach is intended to minimize this effect.

Digital Twin Validation of Photonic Fluid Flow Prediction

This simulation benchmarked the photonic model against a classical reservoir computing model, revealing performance differences and estimating the scale at which photonic simulation would exceed available computational resources. The team evaluated the system’s ability to predict two distinct fluid flow scenarios, beginning with a two-dimensional thermal convection flow described by a truncated Galerkin model with eight degrees of freedom, termed the L8 model. Further investigation assessed the robustness of the photonic model against photon loss, simulating rates up to 50% to reflect real-world limitations of photonic reservoir computing devices.

BSBS Performance Compared to Coherent Light

Simulations reveal that the bright scattershot boson sampling (BSBS) approach consistently outperforms a classical analog using coherent light when modeling complex fluid dynamics. The team’s digital twin of the Paderborn Quantum Sampler, a photonic quantum computer, accurately reproduced the temporal dynamics of both an eight-dimensional thermally driven convection model and the atmospheric circulation model. This superior performance extends to resilience against signal degradation; the photonic model maintained accuracy even with simulated photon losses up to 50%.

Reservoir Computing Model Mapped onto BSBS Architecture

Researchers evaluated the photonic reservoir computing model against a classical counterpart, revealing a performance advantage, particularly when modeling the higher-dimensional Lorenz 96 system representing atmospheric circulation. The architecture’s efficacy extends to scenarios with signal degradation. Simulations accounted for photon loss, showing the BSBS model maintained accuracy even with up to 50% loss of photons.

The researchers found that the photonic model’s prediction horizon exceeded that of the classical reservoir-computing model, suggesting a potential for longer-term forecasting capabilities. This approach offers a novel pathway for using photonic quantum computers not merely for computation, but for modeling and understanding complex physical systems.

Lorenz and Atmospheric Circulation Models Demonstrated

A digital twin, built using multiphoton statistics, confirmed the photonic approach’s ability to reproduce the temporal dynamics of these systems even with up to 50% photon loss, a critical factor for practical implementation. This resilience stems from the system’s design, which allows it to maintain accuracy despite inherent challenges in photonic quantum computing. Configurations varied in interferometer modes, photon sources, vacuum entries, maximum photon number moment order and the number of observables, each impacting the modeling of the Lorenz systems.

Resilience of BSBS to Photon Loss in Simulations

This resilience suggests that the BSBS architecture can maintain accurate temporal dynamics despite inherent challenges in real-world photonic quantum computing. Comparisons between BSBS and runs using coherent light further illustrate the advantages of the quantum approach, particularly for the higher-dimensional Lorenz 96 model. The researchers state that “An application case that is sufficiently high-dimensional outperforms the corresponding classical algorithm even with included errors,” suggesting a promising route toward photonic quantum advantage. This capability extends beyond mere accuracy, indicating that the BSBS method offers a robust pathway for using photonic quantum computers in complex modeling applications.

👉 More information
🗞 Complex fluid flow dynamics by bright scattershot boson sampling
✍️ Robert Schade and Christian Plessl (Paderborn Center for Parallel Computing, Paderborn University, Paderborn, Germany.); Shailendra K. Rathor and Jörg Schumacher (Institute of Thermodynamics and Fluid Mechanics, Technische Universität Ilmenau, Ilmenau, Germany); Tim J. Bartley, Benjamin Brecht, Jan Sperling, Michael Stefszky and Christine Silberhorn (Institute for Photonic Quantum Systems (PhoQS), Paderborn University, Paderborn, Germany)
🧠 ArXiv: https://arxiv.org/abs/2611.18510v1

Stay current

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

Avatar of Dr. Donovan

Latest Posts by Dr. Donovan: