Adiabatic Reverse Annealing Succeeds in Open Quantum Systems, Demonstrating Robustness to Low-Temperature Decoherence

The challenge of finding the lowest energy state of complex systems drives research in diverse fields, and adiabatic reverse annealing (ARA) represents a promising technique for tackling this problem, offering potential advantages over conventional quantum annealing. An Le and Christopher L. Baldwin, from Michigan State University, investigate the robustness of ARA in realistic conditions where environmental noise causes decoherence, a major obstacle to quantum computation. Their work addresses recent suggestions that ARA’s benefits disappear when decoherence is present, and instead demonstrates that ARA can succeed in open systems, provided the temperature of the environment remains sufficiently low. By employing a solvable model and the adiabatic master equation, the researchers reveal that environmental temperature plays a critical role, and surprisingly, can even enhance the performance of ARA under certain conditions, opening new avenues for practical quantum optimisation.

Time Evolution of Magnetization with Noise and Fields

This research details a comprehensive derivation of equations describing the behavior of magnetization within the p-spin model, incorporating the effects of time-dependent external fields and environmental noise. The work culminates in a numerical algorithm for simulating the system’s evolution, providing a detailed understanding of its dynamics. This framework accurately models the magnetization’s response to both external control and the inherent randomness of its environment. The team employed a master equation approach to describe how the density matrix of each spin evolves over time, breaking it down into components representing coherent evolution, energy level shifts, and the effects of dissipation and dephasing.

Jump operators modeled transitions between energy levels due to environmental interactions, and the strength of the noise was characterized by a bath correlation function. The resulting equations were then used to develop a numerical algorithm that iteratively updates the density matrices of the spins, allowing researchers to calculate the magnetization as a function of time. The derivation assumes weak coupling between the spin and the environment, and that the external fields change slowly over time. The algorithm is self-consistent, meaning the magnetization at any given time depends on its previous state, allowing for a numerical solution to the complex dynamics of the p-spin model under various conditions.

Adiabatic Annealing Performance Under Decoherence Effects

Researchers investigated the performance of adiabatic reverse annealing (ARA), an enhanced quantum annealing method, in realistic conditions where systems interact with their environment. Recognizing that ARA’s advantages might diminish with environmental interference, the study used a theoretical approach, employing the p-spin model as a solvable example to analyze ARA’s success under the influence of decoherence. The core of the methodology involved demonstrating that, under adiabatic conditions, the system follows the instantaneous equilibrium state, provided the process avoids finite-temperature phase transitions. Researchers identified two primary mechanisms by which ARA can fail at higher temperatures: either no transition-avoiding paths exist, or the equilibrium state itself becomes disordered.

The team established that sufficiently low temperatures are crucial for ARA’s success, ensuring that neither of these failure mechanisms occur. Remarkably, the study revealed scenarios where the environment actually benefits ARA, finding parameter values where transition-avoiding paths are impossible at zero temperature but emerge at non-zero temperatures. To explore this, the team investigated the finite-temperature phase diagram of a spin-boson model.

Adiabatic Annealing Succeeds With Realistic Noise

Scientists have demonstrated that adiabatic reverse annealing (ARA), an improvement over conventional quantum annealing, can succeed even in open quantum systems where environmental noise is present, provided the temperature of the environment remains sufficiently low. This work offers an analytical perspective on the process, addressing recent numerical studies suggesting ARA loses its advantage in the presence of decoherence. The study confirms that in the adiabatic limit, where the annealing process is sufficiently slow, the system follows the instantaneous equilibrium state determined by the environment’s temperature. The team found that ARA succeeds when the temperature is low enough to avoid phase transitions, either because no transition-avoiding paths exist or because the equilibrium state itself is disordered. Remarkably, the research reveals that a low, but non-zero, temperature can actually improve ARA’s performance, enabling it to circumvent phase transitions that would otherwise be insurmountable at zero temperature. Specifically, the team determined that the success of ARA hinges on the existence of paths in the finite-temperature phase diagram of the model that avoid discontinuous transitions.

Adiabatic Annealing Succeeds With Realistic Noise

This research demonstrates that the adiabatic reverse (ARA) approach to quantum annealing can succeed even in realistic, imperfect quantum computing environments. While previous numerical studies suggested ARA’s benefits diminish with environmental noise, this work establishes analytical conditions under which ARA can effectively locate optimal solutions. The team showed that, provided the environmental temperature remains sufficiently low, ARA successfully follows the instantaneous equilibrium state of the system, circumventing problematic phase transitions. Importantly, the findings reveal a surprising benefit of a low but non-zero temperature; in certain scenarios, a warmer environment actually improves ARA’s performance, enabling it to avoid transitions that would otherwise be insurmountable at zero temperature. This success hinges on the existence of transition-avoiding paths within the system’s phase diagram and the ability to reach an equilibrium state close to the desired ground state. These findings demonstrate that careful control of the environment’s temperature can unlock the potential of ARA for solving complex optimization problems.