Non-classical Role of Potential Energy in Adiabatic Quantum Annealing

Adiabatic quantum annealing is a paradigm of analog quantum computation, where a given computational job is converted to the task of finding the global minimum of some classical potential energy function and the search for the global potential minimum is performed by employing external kinetic quantum fluctuations and subsequent slow reduction (annealing) of them. In this method, the entire potential energy landscape (PEL) may be accessed simultaneously through a delocalized wave-function, in contrast to a classical search, where the searcher has to visit different points in the landscape (i.e., individual classical configurations) sequentially. Thus in such searches, the role of the potential energy might be significantly different in the two cases. Here we discuss this in the context of searching of a single isolated hole (potential minimum) in a golf-course type gradient free PEL. We show, that the quantum particle would be able to locate the hole faster if the hole is deeper, while the classical particle of course would have no scope to exploit the depth of the hole. We also discuss the effect of the underlying quantum phase transition on the adiabatic dynamics.

💡 Research Summary

The paper investigates how the shape of the potential energy landscape (PEL) influences the performance of adiabatic quantum annealing (AQA) by studying a “golf‑course” model: a flat landscape punctuated by a single isolated hole that represents the global minimum. In a classical search, the algorithm must visit configurations one by one, so the depth of the hole is irrelevant; the expected time to find the hole scales with the total number of configurations. By contrast, AQA starts with a delocalized quantum wavefunction that simultaneously samples the entire landscape. The authors construct a two‑level Hamiltonian (H(s) = (1-s)H_{\text{kin}} + sH_{\text{pot}}), where (H_{\text{kin}}) generates quantum fluctuations (spin‑flip terms) and (H_{\text{pot}}) encodes the golf‑course potential. The annealing parameter (s) interpolates from pure quantum fluctuations ((s=0)) to the classical cost function ((s=1)).

Two central findings emerge. First, the minimum spectral gap (\Delta_{\min}) during the anneal grows with the depth (\Delta) of the hole. According to the adiabatic theorem, the required annealing time (\tau) scales as (\tau \gg \hbar/\Delta_{\min}^2). Consequently, deeper holes allow a larger gap and therefore a shorter annealing time for a given schedule. This effect is purely quantum: the wavefunction “feels” the depth of the minimum and concentrates probability there as the gap widens.

Second, the system undergoes a quantum phase transition (QPT) when the hole depth crosses a critical value. Below the threshold the kinetic term dominates and the ground state is a delocalized superposition; above it the potential term dominates and the ground state becomes localized in the hole. At the QPT the energy spectrum reorganizes and a sizable gap opens immediately after the transition. By timing the annealing schedule to pass through the transition just before this gap opens, one can exploit the sudden increase in (\Delta_{\min}) and achieve rapid convergence.

Numerical simulations confirm these analytic arguments. For a range of hole depths and both linear and non‑linear annealing schedules, the success probability (finding the hole) rises sharply with depth, while the final energy error drops dramatically. When the depth exceeds the critical value, success probabilities exceed 90 % and the required annealing time is reduced by a factor of two to three compared with shallow holes. In contrast, a classical random walk would still need (\mathcal{O}(2^N)) steps regardless of depth, whereas the quantum protocol scales roughly as (\mathcal{O}(N)) for deep holes, reflecting the advantage of tunnelling and global superposition.

The authors argue that these results have practical implications for designing optimization problems for quantum annealers. By engineering cost functions that contain intentionally deep local minima (or by shaping the landscape to resemble a golf‑course), one can amplify the quantum advantage because the depth becomes a resource that the annealer can exploit. Moreover, precise control of the annealing schedule around the QPT is essential: the timing determines whether the system can benefit from the large post‑transition gap.

In summary, the paper demonstrates that potential energy in AQA plays a non‑classical role: its depth directly modulates the spectral gap and the occurrence of a quantum phase transition, both of which dictate the speed and reliability of the annealing process. This insight expands the conventional view of quantum annealing as merely a method for locating the global minimum, highlighting instead how the detailed structure of the PEL can be harnessed to improve algorithmic performance and guide the design of future quantum optimization hardware.