This study presents a new method for simulating time evolution in quantum systems, focusing on many-body interactions. Despite the importance of tackling these complex systems in modern physics, their computational complexity has remained a significant obstacle for several decades, making exact solutions scarce, especially when involving dozens of particles or more. Current technological advancements cannot keep pace with this exponential increase in complexity for solutions of a many-body quantum system via classical methods.
Throughout their research, the team employed both tensor network techniques and quantum Monte Carlo methods for studies of specific quantum systems and in certain instances of ergodicity. However, each strategy has its limits, with the first commonly encountering the ‘entanglement barrier’, limiting its capabilities with long-time dynamics. Another challenge that arises stems from computational burden originating in the generation of entanglement, particularly for two-dimensional systems that require substantial bond dimensions with large complexity. Meanwhile, quantum Monte Carlo approaches suffer from inherent difficulties such as the notorious ‘sign problem.’
Aiming towards the creation of flexible and dependable quantum simulators, researchers investigated a novel solution – the use of scrambling transforms in conjunction with continuous unitary transforms, frequently referred to as flow equations. Combined, the strategies enable effective approximation of diagonalizing large Hamiltonian matrices, improving the overall computation of time evolution.
Utilizing this new technique, the research illustrates impressive results in investigations of quantum systems with interacting fermions, demonstrating impressive accuracy and longevity on time scales much longer than classical methodologies traditionally accommodate in such quantum systems. With substantial improvements and technological advancements on horizon, the computational techniques devised through this study provide tantalizing hints to progress yet to be attained through classic computers when simulating various fascinating quantum matters.