The parametric instability landscape of coupled Kerr parametric oscillators
Orjan Ameye, Alexander Eichler, Oded Zilberberg · 2025
Read The parametric instability landscape of coupled Kerr parametric oscillators by Orjan Ameye, Alexander Eichler, Oded Zilberberg (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-08793.txt.
Introduction
Orjan Ameye and co-authors map the stability of coupled oscillators in a 2025 physics paper. The devices, called Kerr parametric oscillators, are driven circuits that can settle into distinct states, and networks of them are candidates for new kinds of computing. Their tipping points, though, form a rich pattern that was largely uncharted.
Starting with two coupled units, the team used perturbation theory to chart where the system stays stable and where it flips. The flips arise from a contest between the shared drive and the coupling between units. Extending the math to large, evenly coupled networks, they found the flip points form a cascade that becomes evenly spaced as the network grows. The results mark the exact bounds where such networks behave like an Ising model, the workhorse of computing with spins, giving builders a guide for experiments.
Catalog record
| CREATOR | Orjan Ameye, Alexander Eichler, Oded Zilberberg |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 75 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-08793
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
The parametric instability landscape of coupled Kerr parametric oscillators Orjan Ameye, Alexander Eichler, Oded Zilberberg Abstract Networks of coupled Kerr parametric oscillators (KPOs) hold promise as for the realization of neuromorphic and quantum computation. Yet, their rich bifurcation structure remains largely not understood. Here, we employ secular perturbation theory to map the stability regions of these networks, and identify the regime where the system can be mapped to an Ising model. Starting with two coupled KPOs, we show how the bifurcations arise from the competition between the global parametric drive and linear coupling between the KPOs. We then extend this framework to larger networks with all-to-all equal coupling, deriving analytical expressions for the full cascade of bifurcation transitions. In the thermodynamic limit, we find that these transitions become uniformly spaced, leading to a highly regular structure. Our results reveal the precise bounds under which KPO networks have an Ising-like solution space, and thus provides crucial guidance for their experimental implementation. I Introduction The Kerr Parametric Oscillator (KPO) stands out as a ubiquitous out-of-equilibrium system in the field of quantum mechanics and nonlinear dynamics [ 1 , 2 ] . This system, driven by a time-dependent harmonic potential term, has emerged as a versatile platform with applications ranging from quantum information processing [ 3 , 4 ] , metrology [ 5 , 6 ] , and ultra-sensitive detection [ 7 , 8 , 9 ] , to neuromorphic computing [ 10 , 11 ] . In quantum information, KPOs serve as qubits and quantum gates, while in sensing applications, they enable precision measurements through parametric amplification. Their capability to form networks with complex dynamics has made them particularly attractive for neuromorphic architectures, where they can implement both computational and memory functions. When the KPO is driven above its parametric driving threshold, the system is forced across a spontaneous period-doubling \mathbb{Z}_{2} symmetry-breaking phase transition [ 12 , 13 , 10 ] , resulting in a “phase state” oscillation solution at half the drive frequency. Intriguingly, a KPO features two such phase states with equal amplitude, but phases separated by an angle \pi . These states can be mapped to the two polarization states of an Ising spin, “up” and “down”. This analogy has sparked interest in KPO networks as a tool to solve NP-hard problems, such as finding the ground state of an Ising Hamiltonian, where different classes of “KPO Ising machines” with bilinear or dissipative coupling emerged [ 14 , 15 , 16 , 17 , 10 , 18 , 19 , 20 , 21 , 22 , 23 ] . The exploration of coupled KPOs thus opens up new possibilities in the realm of computation and optimization, making it a captivating subject of research. A key challenge for using KPO networks as Ising machines is to understand the plethora of phases in the driven many-body system, and how they map to a desired Ising Hamiltonian. Previous studies on networks with bilinear coupling have focused mainly on the case of weak coupling coefficients, where the bare modes of the system can be readily mapped to Ising spin configurations [ 10 ] . In contrast, a strong bilinear coupling between KPOs was found to give rise to additional phases beyond the Ising model [ 20 , 24 ] . In particular, it became clear that the number of maximum stable configurations of an N -node KPO network deviates from the expected solution space of the 2^{N} Ising states. This discrepancy arises on the one hand because the resonator network forms normal modes that are separated in frequency [ 25 ] , and on the other hand because the interplay between nonlinearity and resonator coupling generates a host of additional states with mixed symmetry [ 20 ] . Even for only two KPOs, the combination of these effects results in a complex phase diagram that so far is only partially understood. As such, a deep analytical understanding of the phases associated with two coupled KPOs is crucial for predicting the behavior of larger networks, and eventually for understanding the rules governing large-scale Ising machines. In this work, we present a comprehensive analysis of the phase diagrams of networks of coupled bilinear Kerr parametric oscillators (KPOs). We use secular perturbation theory to understand the stability of the different states in the system, focusing on the interplay between parametric driving and coupling. By examining the case of two coupled KPOs, we derive the conditions under which different stationary states emerge and the physical mechanisms behind them. We extend our analysis to larger networks with all-to-all equally coupling, finding analytical expressions for the bifurcation lines that demarcate the parameter regime where the network can effectively function as an Ising machine. Our findings offer a detailed understanding of the dynamics and solution space of KPO networks, paving the way for their practical implementation in neuromorphic computing applications. The paper is structured as follows: In Sec. II , we provide a comprehensive introduction to linearly coupled KPO networks, where we highlight how cross-Kerr coupling [cf. Eq. ( 12 )] due to strong coupling leads to a rich bifurcation structure. We proceed in Sec. III , by analyzing the stability phase diagram of two coupled KPOs, deriving analytical expressions in Eq. ( 16 ) for the bifurcation lines that determine when different states become stable. Crucially, the stability of the states is perturbed by the coupling between the KPOs. In Sec. IV , we extend this result to networks of N identical all-to-all coupled KPOs, thus revealing a cascade of bifurcation transitions, cf. Eq. ( 23 ). We find that the bifurcation lines ( 27 ) and ( 18 ) define lower bounds on where the system has 2^{N} possible stationary states, thus defining the parameter regime where the system can be mapped to an Ising network. We conclude with a summary and outlook in Sec. V . II Coupled Kerr Parametric oscillators We analyze a network of N coupled KPOs, see Fig. 1 (a). Its classical equations of motion (EOMs) read \ddot{x}_{i}+\omega_{i}^{2}(t)x_{i}+\gamma_{i}\dot{x}_{i}+\alpha_{i}x_{i}^{3}-\sum_{j\neq i}J_{ij}x_{j}=0\,, (1) where the i^{\rm th} EOM describes an KPO by its displacement x_{i} , modulated bare frequency \omega_{i}(t)=\omega_{i}(1-\tilde{\lambda_{i}}\cos(2\omega t)) , and Kerr nonlinearity \alpha_{i} . We work in units such that the mass of each KPO is one, m_{i}=1 . The dots indicate derivatives with respect to time t . All KPOs are parametrically driven at the same angular frequency 2\omega with parametric pump strengths \tilde{\lambda}_{i}\equiv\lambda_{i}/\omega_{i}^{2}\equiv\lambda/\omega^{2}_{i} . The network is formed by linear coupling between the KPOs with coefficients J_{ij} . To account for dissipation in the system, we introduced linear damping terms with coefficients \gamma_{i} . For completeness, we present our results also using equivalent quantum notation by quantizing the phase space coordinates x_{i} and p_{i}\equiv\dot{x}_{i} in Eq. ( 1 ). Namely, we employ raising and lowering operators \hat{x}_{i}\equiv\sqrt{\hbar/2\omega}\,(\hat{a}_{i}^{\dagger}+\hat{a}_{i}) and \hat{p}_{i}\equiv\sqrt{\hbar\omega/2}\,(\hat{a}_{i}^{\dagger}-\hat{a}_{i}) . The resulting Hamiltonian (for \gamma_{i}=0 ) reads \frac{H}{\hbar}=\sum_{i}\frac{\omega_{i}^{2}(t)+\omega^{2}}{2\omega}\hat{a}_{i}^{\dagger}\hat{a}_{i}+\frac{\omega_{i}^{2}(t)-\omega^{2}}{4\omega}(\hat{a}_{i}^{\dagger}\hat{a}_{i}^{\dagger}+\hat{a}_{i}\hat{a}_{i})\\ +\frac{U_{i}}{12}(\hat{a}_{i}^{\dagger}+\hat{a}_{i})^{4}+\frac{1}{2}\sum_{j\neq i}T_{ij}(\hat{a}_{i}^{\dagger}+\hat{a}_{i})(\hat{a}_{j}^{\dagger}+\hat{a}_{j}) (2) where U_{i}=3\alpha_{i}\hbar/4\omega^{2} and T_{ij}=J_{ij}/2\omega . Here, we quantized the system in ladder operators at the frequency of the drive \omega , anticipating the frequency of the response of the oscillators [ 26 , 27 ] . II.1 The linear regime We first recall the linear limit ( \alpha_{i}=0 ) of a