Spin Simulations Now Scale with System Size, Not Exponentially

A new method simulates the complex behaviour of quantum spin systems. Rishab Dutta and colleagues at University of Washington present the fermionized time-dependent Hartree-Fock (fTDHF) approach, a technique for modelling real-time quantum dynamics of spin-1/2 Hamiltonians. The method efficiently addresses challenges posed by long-range interactions, offering a computationally scalable alternative to exact dynamics with a cost that increases polynomially with system size. Benchmarking against three distinct spin models, adiabatic state preparation, disorder-driven localisation, and particle production, demonstrates fTDHF’s ability to accurately reproduce qualitative dynamics while retaining a clear physical interpretation.

Fermionisation of spin interactions enables efficient simulation of complex quantum dynamics

A five-fold improvement in simulating quantum spin systems has been achieved, reducing computational cost from scaling with the cube of system size to polynomially with it. Enabled by the fermionized time-dependent Hartree-Fock (fTDHF) method, this breakthrough unlocks the ability to model systems previously intractable due to exponential computational demands. Accurately simulating even moderately sized spin systems previously required prohibitive computational resources, limiting our understanding of collective quantum phenomena. The method transforms interactions between spins into interactions between fermions, simplifying calculations while retaining important physical accuracy, and was successfully benchmarked against three distinct spin models exhibiting complex behaviours. This transformation is achieved via the Jordan-Wigner transformation, a mapping that represents spin operators in terms of fermionic creation and annihilation operators.

These models, adiabatic state preparation, disorder-driven many-body localisation, and particle production, demonstrate its flexible nature and reliability in capturing qualitative dynamics. Fermionized time-dependent Hartree-Fock (fTDHF) efficacy was demonstrated by benchmarking against exact dynamics for spin-1/2 models. These models represented adiabatic state preparation with long-range correlations, disorder-driven many-body localisation, and particle production in the Schwinger model. Adiabatic state preparation involves slowly evolving a system to a desired ground state, while disorder-driven many-body localisation describes the transition from a conducting to an insulating phase due to strong disorder. The Schwinger model, a quantum electrodynamics analogue, allows investigation of particle creation from the vacuum.

For each system, fTDHF reproduced the qualitative dynamics generated by exact evolutions, maintaining a simple physical picture due to its mean-field nature. Applying the method to a system with 13 spins, researchers confirmed its accuracy by examining correlations closely matching those obtained via exact time evolution. Specifically, they analysed spin-spin correlation functions to quantify the agreement between fTDHF and exact results. Furthermore, fTDHF closely tracked exact results at higher disorder levels when simulating a ten-spin system with varying disorder strengths, consistent with expectations from perturbation theory. This consistency suggests that fTDHF captures the essential physics even in strongly disordered regimes. The approach can be implemented on a classical computer with a computational cost that scales polynomially with system size and linearly with time steps. This polynomial scaling, approximately O(N³) where N is the number of spins, is a significant advantage over exact diagonalisation methods which scale exponentially, approximately O(2^N).

Scalable quantum dynamics through polynomial complexity reduction

Simulating quantum systems is vital for progress in materials science and fundamental physics, offering insights into phenomena beyond the reach of traditional experiments. Understanding the behaviour of strongly correlated materials, designing novel quantum technologies, and probing the fundamental laws of nature all rely on accurate quantum simulations. However, accurately modelling these systems remains computationally demanding, often requiring approximations that sacrifice precision for speed. The fermionized time-dependent Hartree-Fock method represents a step towards bridging this gap, providing a scalable approach to real-time dynamics, although its reliance on a mean-field approximation introduces a fundamental limitation. The mean-field approximation simplifies the many-body problem by replacing interactions between particles with an average field, neglecting correlations.

This scalability allows investigation of larger, more realistic systems previously inaccessible to detailed analysis, opening avenues for exploring phenomena like many-body localisation and particle creation. A computationally efficient way to model the dynamic behaviour of interacting quantum spin systems is now available. By transforming spins into fermions and utilising a mean-field approach, fTDHF circumvents the limitations of previous techniques which struggled with complex, long-range interactions. The Jordan-Wigner transformation is key to this process. The Jordan-Wigner transformation maps spin operators to fermionic operators, allowing the use of established techniques from quantum field theory. This advancement enables exploration of systems with a computational cost that increases at a manageable rate, unlike methods demanding exponentially more resources. This allows focus on refining the approximation and extending its applicability to more complex scenarios. The ability to efficiently handle non-local string operators, arising from long-range interactions, is particularly noteworthy, as these terms often pose a significant challenge for other methods.

The fTDHF method is formally equivalent to exact dynamics in the case of free fermions, providing a rigorous foundation for its accuracy. However, for interacting systems, the mean-field approximation introduces errors, which are expected to be smaller for systems with weak interactions or at high temperatures. Future work could focus on incorporating higher-order correlations to improve the accuracy of the method, potentially through the inclusion of configuration interaction or coupled cluster techniques. The current implementation, while demonstrating significant improvements in computational efficiency, remains limited by the computational cost of solving the time-dependent Hartree-Fock equations for the fermionic system. Nevertheless, the fTDHF method provides a valuable tool for exploring the dynamics of quantum spin systems and offers a promising pathway towards simulating larger and more complex systems in the future. The method’s ability to accurately capture qualitative dynamics, combined with its polynomial scaling, positions it as a significant advancement in the field of quantum simulation.

The researchers developed a new method, fermionized time-dependent Hartree-Fock (fTDHF), for modelling the real-time behaviour of spin-1/2 systems. By transforming spin interactions into a fermionic framework, fTDHF offers a computationally efficient alternative to methods struggling with long-range interactions, scaling polynomially with system size. The method successfully reproduced the qualitative dynamics observed in three benchmark spin models, including those exhibiting many-body localisation and particle production. The authors suggest future work may focus on improving accuracy by incorporating higher-order correlations into the calculations.