Quantum Simulations Become Faster with New Computational Technique

Researchers at the University of St Andrews, led by Elliot W. Lloyd, have developed a new method to simulate the quantum dynamics of complex systems containing numerous interacting emitters. The technique employs a stochastic unraveling approach that respects weak permutation symmetry, substantially reducing computational cost for both two- and d-level emitters. This advancement lowers the computational complexity from $\mathcal{O}(N^5)$ to $\mathcal{O}(N^2)$ for two-level systems, and with additional refinements, allows reduction to $\mathcal{O}(N)$. Consequently, simulations of larger system sizes become feasible, extending the range of accessible quantum phenomena and representing a key step forward in modelling complex quantum behaviours. The approach is not limited to two-level systems; it extends to d-level emitters, scaling as $\mathcal{O}(N^{d(d-1)/2})$, and successfully enables large-N simulations for d=3, demonstrating its versatility and broad applicability to diverse quantum models.

Quantum simulation complexity reduced to linear scaling for two-level systems

The computational burden associated with simulating the quantum dynamics of many-body systems has historically been a significant limitation. This new method addresses this challenge by reducing the scaling from $\mathcal{O}(N^5)$ to $\mathcal{O}(N^2)$, and with further optimisation, achieving $\mathcal{O}(N)$ scaling for two-level emitters. This breakthrough permits the modelling of systems previously considered intractable due to computational constraints, opening new avenues for exploring complex quantum phenomena in areas such as quantum optics and condensed matter physics. The method’s extension beyond two-level systems to encompass d-level emitters, with a scaling of $\mathcal{O}(N^{d(d-1)/2})$, further highlights its flexible and broad applicability to diverse quantum models. Successfully enabling large-N simulations for d=3 level systems demonstrates the practical impact of this development, allowing researchers to investigate systems with a greater degree of complexity than previously possible.

Simulations utilising this technique have successfully modelled systems with three energy levels, exhibiting a scaling behaviour of $\mathcal{O}(N^{3(3-1)/2})$, which simplifies to $\mathcal{O}(N^3)$. Large-N simulations, where N represents the number of interacting components, were achieved for d=3 level systems, a feat previously unattainable with conventional computational resources. The team rigorously verified the method’s accuracy by replicating known results for the Dicke model, a fundamental model in quantum optics, extending simulations to system sizes two orders of magnitude larger than those achieved in prior work. Applying the technique to the Tavis-Cummings model, a variation incorporating conserved excitation numbers, enabled further optimisation by effectively eliminating the need to calculate cavity state contributions between quantum jumps, thereby improving computational efficiency. However, the simulations still require a decreasing timestep as system size increases, a consequence of the inherent dissipation effects within the modelled systems. This suggests that the observed linear scaling may not hold indefinitely at extremely large N, and further algorithmic refinements may be necessary to maintain efficiency at even larger scales.

Stochastic unraveling of permutation symmetric density matrices for many-body emitter simulations

These simulations are predicated on a ‘stochastic unraveling’ technique, which reimagines the complex quantum process as a series of random, individual events. The overall system behaviour is then reconstructed by accurately combining these individual events. This approach directly addresses a common problem in quantum simulation, where computational demands increase exponentially with the number of interacting components. Crucially, the method preserves weak permutation symmetry, a fundamental principle stating that swapping identical particles does not alter the final outcome, analogous to the result of flipping identical coins. This symmetry is a key feature of many physical systems and its preservation is vital for accurate modelling.

By formulating the simulation within a space of permutation symmetric density matrices, a mathematical description of the system’s quantum state, the computational bottlenecks of previous methods were circumvented. A density matrix provides a complete description of the quantum state of a system, including both pure and mixed states. This formulation was strategically chosen to overcome the exponential increase in computational demands associated with conventional methods when dealing with individual dissipation within the system, such as spontaneous emission from the emitters. Dissipation represents the loss of energy from the system to the environment, and accurately modelling this process is essential for realistic simulations. The technique provides a significant advantage over traditional methods, allowing for more efficient calculations by reducing the dimensionality of the problem and exploiting the inherent symmetries of the system. The use of stochastic unraveling allows for the generation of numerous possible trajectories, each representing a single realisation of the quantum process, and averaging over these trajectories provides an accurate approximation of the system’s behaviour.

Reduced computational scaling unlocks more realistic quantum modelling

Simulating quantum systems is becoming increasingly important as physicists strive to model increasingly complex materials and interactions, with applications ranging from the development of new quantum technologies to the understanding of fundamental physical phenomena. This new method offers a pathway to explore previously inaccessible regimes of quantum behaviour, particularly for systems with many interacting components. The abstract acknowledges a potential limitation, however; while successful for three-level systems, computational demands still increase with increasing complexity. Scaling to systems with a significantly larger number of energy levels, or a greater number of emitters, may prove challenging, necessitating further optimisation to prevent prohibitive computational costs. Exploring alternative numerical techniques and leveraging the power of high-performance computing will be crucial for extending the method’s applicability to even more complex systems.

Nevertheless, this advance remains significant for the field of quantum simulation, reducing computational cost from a fifth-power to a quadratic relationship with the number of emitters, and even linearly in certain cases. This opens doors to modelling larger, more realistic quantum dynamics, enabling investigations into phenomena that were previously beyond reach. The new ‘unravelling’ technique simplifies calculations while preserving key quantum behaviours, allowing for the modelling of quantum behaviour in many interacting components with reduced computational cost. This broadens the scope of accessible quantum phenomena and enables investigation of more complex quantum dynamics, in scenarios where computational limitations previously hindered progress. By preserving weak permutation symmetry, ensuring identical particles are treated equivalently, the method achieves a complexity scaling of $\mathcal{O}(N^2)$ for two-level systems, a substantial reduction from $\mathcal{O}(N^5)$, with further refinements reaching $\mathcal{O}(N)$. Extending beyond simpler systems, the technique successfully simulates d-level emitters, allowing large-scale modelling of systems with three energy levels and paving the way for investigations into more complex multi-level quantum systems.

The researchers demonstrated a new technique to model the quantum dynamics of multiple interacting emitters, reducing computational cost significantly. This is important because simulating quantum systems becomes exponentially more difficult as the number of components increases, previously limiting the size and complexity of models. The method achieved a complexity scaling of $\mathcal{O}(N^2)$ for two-level systems, a substantial reduction from $\mathcal{O}(N^5)$, and successfully simulated systems with three energy levels. The authors suggest further optimisation may be needed to extend the technique to even more complex systems with a greater number of emitters or energy levels.