WorksPhysics and Astronomy
Vector Ising Spin Annealer for Minimizing Ising Hamiltonians
Abstract
We introduce the Vector Ising Spin Annealer (VISA), a framework in gain-based computing that harnesses light-matter interactions to solve complex optimization problems encoded in spin Hamiltonians. Traditional driven-dissipative systems often select excited states due to limitations in spin movement. VISA transcends these constraints by enabling spins to operate in a three-dimensional space, offering a robust solution to minimize Ising Hamiltonians effectively. Our comparative analysis reveals VISA's superior performance over conventional single-dimension spin optimizers, demonstrating its ability to bridge substantial energy barriers in complex landscapes. Through detailed studies on cyclic and random graphs, we show VISA's proficiency in dynamically evolving the energy landscape with time-dependent gain and penalty annealing, illustrating its potential to redefine optimization in physical systems.
Introduction
In pursuing advancements in digital computing, mainly aimed at addressing the complexities inherent in AI and large-scale optimization problems, the inherent limitations of the traditional von Neumann architecture come to the forefront. These limitations are characterized by the escalating costs required for incremental performance improvements, prompting a pivotal shift toward more sustainable computational paradigms [1]. In this evolving landscape, two primary paths have emerged: refining algorithms within the existing computational frameworks and exploring novel hardware paradigms grounded in physics principles, with a notable focus on exploiting light-matter interactions.
Algorithmic enhancements, while valuable, offer incremental improvements. In contrast, the exploration of alternative hardware paradigms holds the promise of a substantial shift, leveraging the fundamental principles of physics—such as minimizing entropy, energy, and dissipation [2]—and integrating quantum phenomena like superposition and entanglement [3]. This innovative approach aims to transcend the conventional computing model’s limitations, tapping into the intrinsic computational potential of physical systems to address complex optimization challenges currently beyond traditional methods’ reach.
Central to this innovative trajectory is the integration of analogue, physics-based algorithms and hardware, which involve translating complex optimization problems into universal spin Hamiltonians, like the classical Ising and XY models [4, 5, 6]. This translation process involves embedding the structure of the problem into the spin Hamiltonian’s coupling strengths and targeting the ground state as the solution, with the physical system tasked with discovering this state. The efficiency and accuracy of this mapping process are vital, as they ensure the scalability of computational efforts, enabling these problems to remain manageable despite increasing complexity [7].
Gain-based computing (GBC) using coupled light-matter emerges as a novel disruptive computing platform distinct from gate-based and quantum or classical annealing approaches. GBC combines optical light manipulations with advancements in laser technology and spatial light modulators, facilitating parallel processing across multiple channels with significant nonlinearities and high-energy efficiency. The operational principle of GBC—increasing pumping power (annealing), followed by symmetry breaking and gradient descent—relies on wave coherence and synchronization to lead the system to a state that minimizes losses naturally. Recent developments in physics-based hardware exploiting GBC principles have showcased diverse technologies. These range from optical parametric oscillator based coherent Ising machines (CIMs) [8, 9, 10, 11], memristors [12], and laser systems [13, 14, 15], to photonic simulators [16, 17], polaritons [5, 6], photon condensates [18], and novel computational architectures like Microsoft’s analogue iterative machine [19] and Toshiba’s simulated bifurcation machine [20]. These platforms have been instrumental in minimizing programmable spin Hamiltonians, demonstrating efficacy across a spectrum of complex optimization challenges, including machine learning [21], financial analytics [22], and biophysical modelling [23, 24].
The Ising Hamiltonian, originating from statistical physics, is a natural model that can be implemented in these platforms. It describes interactions between spins in a lattice, providing a fundamental model for understanding complex systems and can be written as
where represents the spin, and is the external magnetic field. is the interaction strength between the -th and -th spins, and is encoded in the coupling matrix . The Ising model’s exploration of ground states in physical systems parallels the search for minimum cost functions in optimization. At the same time, its energy landscapes mirror the loss landscapes in machine learning, offering a unique perspective on optimization problems.
At the core of the hardware representation of Ising Hamiltonians lies the utilization of scalar soft-spins within soft-spin Ising models, which effectively change energy barriers inherent in classical hard-spin Ising Hamiltonians, . The transition from the hard-spin Ising Hamiltonians to the soft-spin Ising models consists of considering binary spins as signs of continuous variables - amplitudes - that bifurcate from vacuum state guided by the gain increase. While facilitating enhanced problem-solving capabilities, this approach occasionally encounters the obstacle of trajectory trapping within local minima due to the escalating energy barriers as gain increases.
Therefore, the challenge of navigating energy landscapes with numerous local minima, especially during amplitude bifurcation, necessitates an innovative approach to avoid the system being trapped in local minima. In this paper, we introduce the Vector Ising Spin Annealer (VISA), a model that integrates the benefits of multidimensional spin systems and soft-spin gain-based evolution. Unlike traditional single-dimension semi-classical spin models, VISA utilizes three soft modes to represent vector components of an Ising spin, allowing for movement in three dimensions and offering a robust framework for accurately determining ground states in complex optimization problems. Our comparative studies underscore the enhanced performance of VISA compared to established models with scalar spins, such as Hopfield-Tank networks, coherent Ising machines, and the spin-vector Langevin model (SVL). VISA showcases a notable ability to overcome significant energy barriers and effectively navigate through intricate energy landscapes. We illustrate VISA’s effectiveness by tackling Ising Hamiltonian problems across various graph structures. Our examples include analytically solvable 3-regular circulant graphs, more complex circulant graphs, and random graphs where minimizing the Ising Hamiltonian is known to be an NP-hard problem. In our discussions, when we refer to “solving a graph A” or “minimizing a graph A”, we use shorthand for the more detailed process of “finding the global minimum of the Ising Hamiltonian on graph A”. This terminology simplifies our reference to determining the lowest possible energy state for the Ising model applied to a specific graph structure.
Vector Ising Spin Annealer
The VISA model is a semi-classical, three-dimensional soft-spin Ising model. It employs annealing of the loss landscape, symmetry-breaking, bifurcation, gradient descent, and mode selection to drive the system toward the global minimum of the Ising Hamiltonian. Here, we represent Ising soft-spins as continuous vectors in three-dimensional space that are free to move in that space. VISA may be physically realized using amplitudes of a network of coupled optical oscillators. In these optical-based Ising machines, vectors of Ising spins are represented by three soft-spin amplitudes. In total, amplitudes are required to minimize an -spin Ising Hamiltonian. This approach is somewhat complimentary to the recently proposed “dimensionality annealing” where soft Ising amplitudes are considered as coordinates of the multidimensional spins such as XY or Heisenberg spins [25, 26]. While the main idea is to exploit the advantages of the higher dimensionality, VISA aims at the Ising minimization rather than hyper-dimensional spin systems [25, 26].
To formulate the Hamiltonian that is capable of representing the dynamics of the GBC, we incorporate the term that is convex and dominates when the effective gain (loss) is large and negative, the Ising term that is minimized when the gain is large and positive, and a term that aligns the spins at the end of the process so that their direction can be associated with the binary spin . We write, therefore, the VISA Hamiltonian as the sum of three terms , where
with hyperparameter . As the effective gain increases with time from negative (effective losses) to positive (effective gain) values, anneals between a convex function with the minimum at to nonzero amplitudes. coincides with the Ising Hamiltonian when all have unit magnitude when projected along the same direction. Finally, is a penalty term with time-dependent magnitude to enforce collinearity between and at the end of the gain-induced landscape change. At which time, the condition on the amplitudes is imposed by the feedback on the gain realised by [27]. Analogously to CIM operation, as grows from negative to positive values, anneals from the dominant convex term that is minimized at for all , to the minimum of with via symmetry-breaking bifurcation. Concurrently, as increases from to sufficiently large , the soft-spins vectors become collinear. The Ising spins are calculated by projecting the three-dimensional vectors along the axis they have centred around and taking signs of the resultant scalar . At , the target hard-spin Ising Hamiltonian is minimized.
The Hamiltonian is 4-local due to the term. Some optical hardware can directly encode such interactions [28]. At the same time, it is possible to reduce the 4-local Hamiltonian to quadratic Hamiltonian by substituting the product of two variables with a new one while adding the appropriate penalty terms. For instance, the quartic interaction term can be reduced to a triplet by introducing a new variable instead of and the term (known as Rosenberg polynomial [29]) in the Hamiltonian where is the penalty constant. Note that if and only if and is strictly positive otherwise. The reduction of the triplet is done similarly by replacing with a new spin variable and introducing the penalty term in the Hamiltonian [30].
The governing equations represent the gradient descent combined with the temporal change of annealing parameters and as
where in the last two terms, the vector triple product has been invoked. The operation of VISA, therefore, relies on the gradient descent of a gain-driven losses landscape.
We illustrate a toy model of VISA for two dimensions in Figure 1, demonstrating how the energy landscape evolves as the gain and penalty are evolved, and how the minimizers and minima of change during this process.

Figure 1. Energy landscape of components of the two-dimensional VISA Hamiltonian for with spin-spin coupling terms and . The normalized minimizer is and , and therefore in this illustration we choose state vectors while setting . Snapshots of the self-interacting term through the annealing schedule are shown in (a)-(c) for and (a), (b), and (c), illustrating the symmetry-breaking as the gain increases. is shown in (d)-(f) which seek to minimise interaction energy between soft spins; in (f) the minimum is achieved as the angle between vectors and is ; in a more general coupling matrix the vector spins will tend to adjust the angles as to minimise the pairwise interactions. is shown in (g)-(i) for the same gain parameter as in (a)-(c) and (g), (h), and (i). The vector spins are forced to be collinear with the direction that can be identified with the hard Ising spins in (i).
Numerical Evolution of Vector Ising Spin Annealers
We consider two popular models to introduce the concept of a scalar-based GBC against which VISA will be benchmarked. (A) Hopfield-Tank (HT) neural networks [31, 32] have energy landscape given by the Lyapunov function
with nonlinear activation function and real soft-spin variables describing the network state. At any time , the Ising state can be obtained from by associating spins with the sign of the soft-spin variables . The governing equation for HT neural networks is with annealing parameter . This first-order equation can be momentum-enhanced and replaced with the second-order equation
leading to Microsoft analogue iterative machine [19] or Toshiba bifurcation machine [33]. We will use these enhancements below. (B) CIM using the degenerate optical parametric oscillators has an energy function
where , , and represent degenerate optical parametric oscillator quadrature, effective laser pumping power, and coupling strength, respectively. The system evolves as
where is a hyperparameter chosen to maximize solution quality. In the scalar-based GBC models, as the gain increases from negative values (representing effective losses) to large positive values (large effective gain), the amplitudes undergo Hopf bifurcation and reach as , for . At the fixed point, the second term in Eq. (6) dominates, which is the target Ising Hamiltonian scaled by .
The energy landscapes during amplitude bifurcation have many local minima from the increased degrees of freedom of the soft-mode systems. Moreover, the ground state of the soft-mode system may correspond to the excited state of the hard-spin Ising model. At the bifurcation, the system trajectories may get trapped at these minima and could not transition to the correct global minimum at higher gains, especially when global cluster spin flips are required [34]. By enhancing the dimensionality of the energy landscape, VISA may allow the system to evolve from the global minimum of the soft-spin Ising model to the global minimum of the classical hard-spin Ising Hamiltonians, as we now illustrate.
J-Möbius Ladder Graph
First, we examine the Ising Hamiltonian minimization on simple cyclic graphs. These graphs offer analytically solvable benchmarks with distinct and identifiable obstacles in finding ground states. The process of finding Ising ground states on such graphs is significantly influenced by the eigenvalues of the coupling matrix , especially the relationship between the eigenvectors’ component signs and the Ising Hamiltonian’s global minimum [34]. For instance, HT networks adjust spin amplitudes to favour the principle eigenvector component signs [35].
We start by using a Möbius ladder-type graph, a cyclic structure with an even number of vertices arranged in a ring, featuring variable couplings between adjacent and opposite vertices. To explore non-trivial ground states, we introduce equal antiferromagnetic couplings () among nearest neighbors and variable cross-ring antiferromagnetic couplings (). We refer to the Möbius graphs with these couplings as -Möbius graph. Following Ref. [34], we define as the alternating spin state around the ring and as the state with spins alternating except at two opposite ring points with frustrated spins, as depicted in Figure 5(a) and (b). When is even, the energies and principal eigenvalues of and intersect at and , respectively. In the range , ’s largest eigenvalues are smaller than those of , with an eigenvalue gap

Figure 5. -Möbius ladder graphs with circle (blue) and cross-circle (red) couplings for the (a) and (b) configurations. (c) Ground state probability for VISA, SVL, ME-HT, CIM, and BFGS for an -Möbius ladder graph with varying cross-circle couplings . One thousand runs are used to calculate the probability of finding the ground state for each value of . Dashed lines corresponding to (black) and (red) show the values of for which the leading eigenvalue changes and the ground state configuration changes, respectively. For VISA, , , and . For SVL, , , and . For momentum-enhanced HT, , and . For CIM we used with , and . For each value of , optimal values of and are chosen based on sets of preliminary runs in which they are varied. Inset (d) illustrates the structure of Möbius ladder coupling matrices with (white), (blue), (red).
even though represents the ground state with lower energy. The dynamics of semi-classical soft-spin models, including these considerations, are juxtaposed with quantum annealing approaches in the minimization of the Ising Hamiltonian on -Möbius ladder graphs, as detailed in Ref. [34].
The presence of amplitude heterogeneities enables the soft-spin model to acquire and maintain its state achieved at the bifurcation even when it diverges from the classical hard-spin Ising Hamiltonian ground state, potentially complicating the optimization process when . To mitigate this issue, the manifold reduction CIM (MR-CIM) technique was developed, incorporating an additional feedback mechanism to regulate soft spin amplitudes, ensuring they remain close to their mean value [34]. The amplitude adjustment after each update at time step is governed by
where . This adjustment draws the spins nearer to the mean, with the average defined by the squared radius of the quadrature , thus aligning the local and global minima of the soft and hard spin models more closely. Comparative analyses between various CIM modes and quantum annealing have been conducted, demonstrating that, despite quantum annealing’s ability to utilize quantum entanglement and inter-spin correlations to identify ground states, it exhibits heightened sensitivity to diminishing energy gaps near . Consequently, quantum annealing demands longer annealing schedules to accurately determine ground state solutions as nears [34].
Figure 2 illustrates the application of VISA to a -Möbius ladder graph instance characterized by cross-circle couplings where . The continuous spin components experience an Aharonov-Hopf bifurcation during the process, exploiting the minimal energy barriers inherent in soft-spin models. We particularly focus on the intermediary time following the bifurcation, noting that at this time, the spin amplitudes have yet to achieve unit magnitude, and the spins are not collinear. By a later time , the spin vectors stabilize as the minimization of the VISA Hamiltonian progresses. These soft-spin vectors then exhibit Ising spin characteristics, including unit magnitude and (anti-)parallel alignment, while the coupling term guarantees identifying the target Ising Hamiltonian’s ground state.

Figure 2. Evolution of the VISA model in an -Möbius ladder network with coupling strength , as governed by Eq. (5). This figure illustrates the dynamics of VISA’s soft-spin components: (a), (b), and (c), with vertical dashed lines indicating the changes at times and . Panels (d) and (e) depict the corresponding spin vectors, highlighting the orientation at —influenced by —and at , where spins align in a unified direction due to a high , revealing the hard-spin Ising Hamiltonian’s global minimum. The final orientation of the vectors at is spontaneous; thus, the vectors are adjusted such that the -axis aligns with this direction. Panel (f) shows the trajectory of the VISA Hamiltonian, , over time as it converges to the ground state. The numerical integration of Eq. (5) utilized the following parameters: , , , , and initial conditions uniformly distributed within . The fourth-order Runge-Kutta method with fixed time step was employed for solving Eq. (5).
In Figure 3, we depict typical trajectories of VISA vectors on Bloch spheres, drawing a parallel to similar quantum representations. We orient the spontaneously aligning spins along the -axis to facilitate visualisation. With a fixed interaction strength where and considering , we demonstrate the bifurcation dynamics originating from the center. This visualization shows how the three-dimensional nature of VISA allows spins to traverse paths that connect various minima, ultimately converging on the ground state . The final vector states are indicated by black arrows, while red points mark the terminal positions of spin vectors at sequential time steps . These trajectories elucidate the process by which spins increasingly satisfy collinearity and unit magnitude constraints through external annealing of and the feedback mechanism on , steering the spin magnitudes to the threshold value of .

Figure 3. Evolution of VISA spins for an -Möbius graph with coupling constant . The spins reach final states adhering to the Ising model’s collinear and unit magnitude constraints. The spontaneous orientation of vectors at time leads to an adjustment of the spin vectors, aligning the -axis with the resultant direction. Parameters used are , , , , and initial conditions uniformly distributed within .
In Figure 4, we compare VISA with CIM under equivalent starting conditions, highlighting that while CIM cannot find the ground state within the range , VISA excels by leveraging the dynamics of spins in three dimensions to identify the global minimum. Soft-spins contribute to this success in both models by facilitating the escape from local minima through reduced energy barriers, a feature not present in hard-spin models.

Figure 4. Panel (a) compares the trajectory evolution of VISA (blue) and CIM (red) on an -Möbius ladder graph with a coupling constant . Here, VISA successfully reaches the ground state , while CIM remains in an excited state . Both systems start from the same initial conditions, with CIM beginning at and VISA at , where and are uniformly chosen from the ranges and respectively. Panels (b)-(d) for CIM and (e)-(g) for VISA show the spin states at three different times , , and , marked by vertical dashed lines in panel (a), illustrating CIM’s inability to overcome the energy barrier to reach the ground state, in contrast to VISA’s effective navigation in three-dimensional space. Panel (h) compares the VISA Hamiltonian against the CIM energy as calculated from Eq. (7), while panel (i) focuses on the Ising energy, with the ground state indicated by a black dashed line. Lastly, panel (j) outlines the regions in the - space that correspond to different global minima of for , showing in blue and in varying colors (yellow, orange, red) for values of 0.35, 0.40, and 0.45, respectively.
els. Unlike CIM, constrained by the principal eigenvalue of to an excited state, VISA uses its multidimensional advantage to bridge minima unreachable by CIM’s one-dimensional approach. Additionally, VISA’s mechanism allows for spin flipping at energy costs lower than that required by CIM, thanks to its ability to navigate through three-dimensional space to find the most energy-efficient path to the global minimum. Figures (4)(a)-(i) depict how VISA effectively minimizes Ising energy, contrasting with CIM’s stagnation in an excited state. For , and as the gain and penalty increase, VISA goes towards the ground state , aligning with the lowest energy state indicated by the hard-spin Ising Hamiltonian, as shown in Figure 4(j). Nonetheless, it’s important to note that increasing spin amplitudes too fast may also inadvertently heighten energy barriers, potentially impeding state transitions. Further insight into the structure of the VISA energy landscape and its comparison with the energy landscape of the scalar models can be gained by analysing the critical points (see Supplementary Information). The VISA energy landscape shows much-diminished energy barriers.
We compare VISA to two second-order scalar networks methods, namely (i) momentum-enhanced Hopfield-Tank (ME-HT), and (ii) spin-vector Langevin. With rigorous hyper-parameter exploration phases, ME-HT outperforms parallel tempering, simulated annealing, and commercial solver Gurobi at various QUBO benchmarks [19]. The ME-HT governing equation is
with effective mass , momentum term , and hyperparameters and , the latter of which undergoes an annealing protocol given by . Equation (10) combines gradient descent with annealed non-conservative dissipation, and the addition of momentum distinguishes it from regular first-order HT networks with energy landscape Eq. (6). Momentum, or the heavy-ball method, aims to overcome the pitfalls of pathological curvature in deep learning and accelerates standard gradient descender optimizers [36, 37].
VISA can be further compared and contrasted with the spin-vector Langevin (SVL) model that was proposed as a classical analogue of quantum annealing description using stochastic Langevin time evolution governed by the fluctuation-dissipation theorem [38]. SVL is based on the time-dependent Hamiltonian used in quantum annealing , where initial Hamiltonian , and problem Hamiltonian , with Pauli operator acting on the -th variable. Real annealing functions satisfy boundary conditions and , where is the temporal length of the annealing schedule. If the rate of change of the functions is slow enough, the system stays in the ground state of the instantaneous Hamiltonian so that the Ising Hamiltonian is minimised at . Quantum annealing has shown competitive results in quadratic unconstrained binary optimization (QUBO) problems such as subset sum, vertex cover, graph coloring, and travelling salesperson [39]. The SVL model replaces Pauli operators with real functions of continuous angle , , and is therefore a classical annealing Hamiltonian using continuous-valued spins . SVL dynamics is described by a system of coupled stochastic equations
where is the effective mass, is the damping constant, is a hyperparameter, and is an iid Gaussian noise. For long annealing times, the minima of HSIHs are obtained through the transformation . The gradient term in Eq. (11) is
which in conjunction with fluctuation-dissipation relations and , give stochastic differential equations: and
where represents a real-valued continuous-time stochastic Wiener process [38]. The characteristic amplitude bifurcation of scalar spins according to ME-HT and SVL are given in Supplementary Information.
The key feature distinguishing VISA from the discussed models lies in its novel gain-based and dimensionality annealing strategy applied across multiple dimensions. In the analysis of -Möbius ladder graphs, as shown in Figure 5, we compute the ground state probability for VISA alongside SVL, ME-HT, CIM, and the Broyden-Fletcher-Goldfarb-Shanno (BFGS) algorithm [40]. Within the range , VISA consistently identifies the ground state with a higher probability compared to the other models.
Cyclic Graphs
Next we consider the cyclic graphs that are a variant of -Möbius ladder graphs that include additional couplings, connecting each vertex to vertices with a weight of , where . This modification aims to explore more complex interaction patterns and their impact on the system’s ground state behavior and eigenvalue distributions, particularly focusing on how these factors evolve in optimization and the search for ground states in varied graph structures. The interaction strength range is expanded to accommodate both ferromagnetic and antiferromagnetic interactions, with . This yields a 5-regular circulant graph characterized by weights , depicted in Figure 6. Cyclic graphs maintain their local and global topological properties under rotational transformations, encapsulating all connectivity information within any row of . By selecting the first row , we compute eigenvalues as , leading to [41]. We then deduce boundaries by observing eigenvalues in the plane, identifying where leading eigenvalues and corresponding eigenvectors change their correspondence with the ground and excited states. The ground state boundaries are identified within the space in the Supplementary Information. Figure 6 depicts regions where the global minimum diverges from the hypercube corner of the projected eigenvector with the highest eigenvalue.

Figure 6. cyclic graphs for with couplings (a) , and (b) . Panel (c) depicts the matrix structure for , where the color coding represents (white), (blue), (red), and (green). The ground states and leading eigenvalue states for are displayed in panels (d) and (e) within the space, respectively. Panel (f) delineates the energy and eigenvalue boundaries, with the pink region indicating discrepancies between ground and maximal eigenvalue states. The sequence from panels (g) to (i) extends this analysis to , highlighting the ground state probabilities across easy-hard transitions in the space, indicated by solid red lines in panels (f) and (i).
We investigate ground state probabilities over transition regions in these graphs with both and cross-circle couplings. We choose values in parameter space that demonstrate transitions between regions in which the eigenvector corresponding to the leading eigenvalue does not match the ground state solution (exact calculations of these regions are presented in Supp. Inf.). Ground state probabilities for VISA, SVL, ME-HT, CIM, and BFGS are illustrated in Figure 7. Four cases are analyzed, split into two sets: and . We consider perpendicular lines in the two-dimensional space for each set. Specifically, we vary (fix) and fix (vary) . For and , an easy-hard-easy transition emerges as increases, akin to -Möbius ladder previously studied.

Figure 7. Ground state probability for VISA, SVL, ME-HT, CIM, and BFGS for cyclic graphs with varying cross-circle couplings and . (a) , , and . (b) , , and . (c) , , and . (d) , , and . One thousand runs are used to calculate the probability of finding the ground state for each value of . Dashed red and black lines show where the ground state energies and leading eigenvalues change, respectively. For VISA, , , and . For SVL, , , and . For momentum-enhanced HT, , and . The general annealing protocol for CIM is used with and . For each value of , optimal values of and are chosen based on sets of preliminary runs in which they are varied.
Indeed, for , decreases as the eigenvalue gap increases. If, instead, we fix , a transition occurs, centred at the change between ground states given by . Here, the hard region is bounded by eigenvalue crossing points and . For , is less sensitive to the magnitude of for regions where is the ground state. This is due to the proximity between the leading eigenvalue state and ground solution in topological spin space. More precisely, the transformation from to requires only a single spin flip, representing a nominal energy barrier for soft-spin models. Therefore, hardness in cyclic graphs derives not just from the eigenvalue gap magnitude but additionally from the distance between hypercube corners of the ground and leading eigenvalue states.
Random Graphs
We extend VISA’s evaluation to QUBO instances renowned for their computational intensity as they scale: dense fully connected graphs and sparse threeregular graphs with customarily distributed random couplings , representing Sherrington-Kirkpatrick (SK) and weighted three-regular Max-Cut problems, respectively. These models are pivotal for benchmarking physical simulators [42, 43, 44, 45] and are categorized under the NP-hard complexity class [4, 46]. Unlike cyclic graphs, these instances do not have analytically known ground states, necessitating using the Gurobi optimization suited for ground state estimations. Gurobi employs advanced pre-processing and heuristic enhancements to expedite branch-and-bound algorithms [47]. Figure 8(a) compares the ground state approximations achieved by Gurobi with those by VISA for SK and weighted three-regular Max-Cut graphs. Alongside, we include comparisons with SVL, MR-CIM, and CIM, where MR-CIM and CIM adapt their laser intensities following a general pumping scheme . MR-CIM further applies an additional feedback mechanism as per Eq. (9) on top of Eq. (8), controlling soft-spin amplitudes and the dimensionality landscape. We define the quality improvement of VISA over another method X in terms of objective values as , showcasing these metrics for SK and three-regular problems in Figure 8(b), where X represents the best-performing competing method for each instance.

Figure 8. (a) The proximity gap, defined as the ratio of the found objective to the Gurobi objective, for VISA, SVL, manifold reduction CIM, and CIM methods. Gurobi was given a 30-second time limit for each of the 100 problem instances with size . The instances are divided equally between two graph topologies: dense, fully connected, and sparse three-regular graphs. In both cases, the matrix weight elements are drawn from the Gaussian distribution with zero mean and unit variance, resulting in instances belonging to the Sherrington-Kirkpatrick and weighted 3-regular Max-Cut problems. (b) Violin plots demonstrate the distribution of VISA’s quality improvement performance compared to the best solution found by competing methods {SVL, MR-CIM, CIM} across the SK and 3-Regular QUBO benchmarks.
Conclusions
This paper introduces the Vector Ising Spin Annealer (VISA). This model capitalizes on the advantages of multidimensional spin systems, gain-based operation and soft-spin annealing techniques to optimize Ising Hamiltonians on various graph structures. VISA distinguishes itself by enabling more efficient navigation through the solution space, enhancing spin mobility in higher-dimensional spaces, and providing a robust framework for connecting local minima and reducing energy barriers.
A key focus of VISA is its ability to recover ground states effectively, even in scenarios where these states do not correspond to the principal eigenvector of the coupling matrix. The model’s performance was numerically evaluated against other methods, demonstrating its superior ability to find ground states across different graph types and complex QUBO instances. Thus, it highlights its potential to address NP-hard problems.
Future research could explore the role of defects in spin models, such as topological defects, domain walls, and vortex rings and their role in achieving the ground state during gain-based operation. Vortices may exhibit more efficient annihilation properties in higher-dimensional systems, such as those utilized by VISA. This is potentially due to the additional spatial degrees of freedom, which could facilitate the merging or cancellation of vortices and anti-vortices, a phenomenon less constrained than in two-dimensional spaces. This enhanced annihilation could lead to a smoother energy landscape, aiding the system in avoiding local minima and more effectively converging to the ground state.
By leveraging multidimensional spins, VISA opens new avenues for developing analogue optimization machines, potentially incorporating quantum effects to enhance computational capabilities. The prospects of applying dimensionality annealing techniques in optical-based Ising machines suggest an exciting future for speed-of-light computation and accurate ground state recovery, marking a significant advancement in optimization technologies.
ACKNOWLEDGEMENTS
J.S.C. acknowledges the PhD support from the EPSRC. N.G.B. acknowledges the support from the Julian Schwinger Foundation Grant No. JSF-19-02-0005, the HORIZON EIC-2022-PATHFINDERCHALLENGES-01 HEISINGBERG project 101114978, and Weizmann-UK Make Connection grant 142568. Both authors are grateful to Dr Fernando Gómez-Ruiz for valuable discussions.
SUPPLEMENTARY INFORMATION
Critical Points of VISA Hamiltonian
To determine the critical points of , we set for all , where ranges from 1 to . Figure 9 depicts these critical points, with the minima corresponding to the states and emphasized. As the gain and the collinearity penalty are systematically increased, a transition in the ground state from to is observed. Notably, there is an infinite continuum of pairs where and share the same energy level. This continuum forms a demarcation line in the space, illustrated in Figure 4(j). At a specific point on this boundary, as depicted in Figure 9(c), both and emerge as ground states. Furthermore, Figure 9 quantifies the average Euclidean distance that each soft-spin needs to traverse to transition from the local minimum to the global minimum , passing through the nearest saddle point. In every scenario analyzed, the VISA model demonstrates a shorter requisite distance compared to the analogous CIM model, underscoring its efficiency in navigating the energy landscape.

Figure 9. Critical points of for -Möbius ladder graph with . Each point is characterized by its energy and Euclidean distance from the origin . The number of soft spins normalizes distance . Blue and red circles denote minima for states and , respectively, while black circles indicate other minima and saddle points of . The closest saddle point (SP) to excited state is highlighted in green. Progressing from figures (a) to (f), we systematically increase the gain for all and the penalty , following a typical annealing schedule. Notably, at (c), with and , we reach a critical juncture that delineates the transition between global minima and , as illustrated in Figure 4(j), where and attain equal energy levels . Adjacent bars to the main plots list distance for journey for VISA (blue) and CIM (red), where for CIM we take .
Basins of Attraction
Figure 10 illustrates the basins of attraction for a range of gain and penalty values on -Möbius ladder graphs with . Panels (a) to (f) in the figure show the evolution of these basins as and are incrementally increased, following a typical gain-based and penalty annealing schedule in the VISA model. The basins of attraction are defined by the initial states , uniformly distributed across the interval , which converge to distinct minima through gradient descent. Initially, at , the global minimum is , which has the largest basin of attraction, capturing all initial states towards . However, as and progressively increase, the attraction basin associated with the excited state expands. Beyond the critical threshold in the parameter space, demonstrated in Figure 4(j), transitions to become the new ground state. Further annealing of and beyond this threshold reveals additional higher-energy states through the process of gradient descent.

Figure 10. Basins of attraction of the VISA model landscape on -Möbius ladder graph. We take and with various gains and penalties as they anneal in the VISA method. Four thousand randomly distributed starting points in are plotted with colors corresponding to minima reached via gradient descent using Newton’s method. (a) , (b) , (c) , (d) , (e) , and (f) . To characterize points, the magnetization magnitude and correlation magnitude are used, where and are the average magnetization and correlations along the -axis, respectively. For small and , the basin of attraction is dominated by the state. As and grow, the volume of the basin of attraction for state increases. For and , which lie on the critical boundary, shown here in (c), both and states have the same energy. As and increase further, becomes the ground state. This is indicated by the switch between excited (red) and ground (blue) states.
Second-Order Scalar Networks Minimizing the Ising Hamiltonian on -Möbius Ladder Graphs
Figure 11 illustrates the Aharonov-Hopf bifurcation in soft-spins within the ME-HT model. For each time step , we apply a clipping function when , alongside a nonlinear activation function . The parameter is adjusted according to the largest positive eigenvalue of the adjacency matrix . Concurrently, Figure 12 displays the temporal evolution of scalar amplitudes as dictated by the SVL model, applied to the same Möbius graph with .

Figure 11. Evolution of momentum enhanced Hopfield-Tank method for an -Möbius ladder network with . (a) The amplitudes connected by the frustrated edges are lower than in the rest of the system and are shown in red. (b) The corresponding Ising Hamiltonian decreases in time, and the ground state is found, with energy given by the dashed red line. (c) Schematic representation of state realised by the soft-spin momentum enhanced HT model. Here, , , , , and is chosen uniformly at random from the range . Fourth order Runge-Kutta is used with fixed time step to solve Eq. (10).

Figure 12. Evolution of the spin-vector Langevin (SVL) model. Applying the SVL model on an -Möbius ladder network with coupling strength . (a) depicts the amplitudes connected by frustrated edges, shown in red, which are lower than those in the rest of the system. (b) illustrates the real annealing functions in blue and in red, satisfying the boundary conditions and , over an annealing period . (c) shows the total Hamiltonian dynamics: it decreases initially as the ground state of is reached, then increases as and are annealed. Post a critical juncture, the SVL model aims to minimize the problem Hamiltonian , identifying the ground state of the -Möbius ladder, indicated by the dashed red line. Parameters used are , , , and noise . Annealing schedules are defined as and . Initial phases are randomly selected from . Equation (11) is solved using a fourth-order Runge-Kutta method with a fixed time step of .
Eigenvalues, Eigenvectors, and Ground States of Cyclic Graphs
Here we refer to cyclic graph defined in the main text, and illustrated in Figure 6(a)-(b) for . For general with couplings , the eigenvalues are given by
where . Equation (14) follows from substituting into the general form of matrix eigenvalues for cyclic graphs given in the main text. As and
vary within , the leading eigenvalue changes. Constrained to this domain, for and , the leading eigenvalues are given by . Substituting these values into Eq. (14) and equating each pair gives three boundaries defined by , , and . Similarly for , we obtain two unique leading eigenvalues in the domain, with boundary . Eigenvectors corresponding to leading eigenvalues are used to deduce analogous Ising states by projecting the eigenvector onto the nearest hypercube corner . These Ising states are given in Figure 6(e) and (h). Gurobi is used to obtain ground states in space. Four unique configurations are found with : namely , , and for , and , , and for . and are defined as the configurations of Ising spins given by and . For even , these four states have energies
where if is even, and 0 otherwise. Equating the relevant energies for gives three boundaries defined by separators , , and . Similarly for , the boundary edges are given by , , and . Eigenvalues and ground states for cyclic graphs are shown in Figure 13 for values of and considered in the main text.

Figure 13. Eigenvalues of cyclic graphs as functions of and . Each plot corresponds to one of the red lines in Figure 6(f) and (i). Ground state probabilities are calculated on these lines in space for various methods in the main text. (a) and . (b) and . (c) and . (d) and . Vertical dashed red and black lines show where ground state energies and leading eigenvalues change, respectively. Background colors denote ground states in specific regions, with states and corresponding colors listed in the key.
Supplementary Information: Vector Ising Spin Annealer for Minimizing Ising Hamiltonians
James S. Cummins; Natalia G. Berloff
Critical Points of VISA Hamiltonian
To determine the critical points of , we set for all , where ranges from 1 to . Figure S1 depicts these critical points, with the minima corresponding to the states and emphasized. As the gain and the collinearity penalty are systematically increased, a transition in the ground state from to is observed. Notably, there is an infinite continuum of pairs where and share the same energy level. This continuum forms a demarcation line in the space, illustrated in Fig. (??)(j). At a specific point on this boundary, as depicted in Figure S1(c), both and emerge as ground states. Furthermore, Figure S1 quantifies the average Euclidean distance that each soft-spin needs to traverse to transition from the local minimum to the global minimum , passing through the nearest saddle point. In every scenario analyzed, the VISA model demonstrates a shorter requisite distance compared to the analogous CIM model, underscoring its efficiency in navigating the energy landscape.

Figure S1. Critical points of for -Möbius ladder graph with . Each point is characterized by its energy and Euclidean distance from the origin . The number of soft spins normalizes distance . Blue and red circles denote minima for states and respectively, while black circles indicate other minima and saddle points of . The closest saddle point (SP) to excited state is highlighted in green. Progressing from figures (a) to (f), we systematically increase the gain for all and the penalty , following a typical annealing schedule. Notably, at (c), with and , we reach a critical juncture that delineates the transition between global minima and , as illustrated in Fig. (??)(j), where and attain equal energy levels . Adjacent bars to the main plots list distance for journey for VISA (blue) and CIM (red), where for CIM we take .
Basins of Attraction
Figure S2 illustrates the basins of attraction for a range of gain and penalty values on -Möbius ladder graphs with . Panels (a) to (f) in the figure show the evolution of these basins as and are incrementally increased, following a typical gain-based and penalty annealing schedule in the VISA model. The basins of attraction are defined by the initial states , uniformly distributed across the interval , which converge to distinct minima through gradient descent. Initially, at , the global minimum is at , which has the largest basin of attraction, capturing all initial states towards . However, as and are progressively increased, the basin of attraction associated with the excited state expands. Beyond the critical threshold in the parameter space, demonstrated in Fig. (??)(j), transitions to become the new ground state. Further annealing of and beyond this threshold reveals additional higher-energy states through the process of gradient descent.

Figure S2. Basins of attraction of the VISA model landscape on -Möbius ladder graph. We take and with various gains and penalties as they anneal in the VISA method. 4000 randomly distributed starting points in are plotted with colors corresponding to minima reached via gradient descent using Newton’s method. (a) , (b) , (c) , (d) , (e) , and (f) . To characterize points, the magnetization magnitude and correlation magnitude are used, where and are the average magnetization and correlations along the -axis, respectively. For small and , the basin of attraction is dominated by the state. As and grow, the volume of the basin of attraction for state increases. For and which lie on the critical boundary, shown here in (c), both and states have the same energy. As and increase further, becomes the ground state. This is indicated by the switch between excited (red) and ground (blue) states.
References
- [1]N. C. Thompson and S. Spanuth, Communications of the ACM 64, 64 (2021).
- [2]S. K. Vadlamani, T. P. Xiao, and E. Yablonovitch, Proceedings of the National Academy of Sciences 117, 26639 (2020).
- [3]E. Farhi, J. Goldstone, S. Gutmann, and M. Sipser, arXiv preprint quant-ph/0001106 (2000).
- [4]A. Lucas, Frontiers in physics 2, 5 (2014).
- [5]N. G. Berloff, M. Silva, K. Kalinin, A. Askitopoulos, J. D. Töpfer, P. Cilibrizzi, W. Langbein, and P. G. Lagoudakis, Nature materials (2017).
- [6]K. P. Kalinin, A. Amo, J. Bloch, and N. G. Berloff, Nanophotonics 9, 4127 (2020).
- [7]G. De las Cuevas and T. S. Cubitt, Science 351, 1180 (2016).
- [8]P. L. McMahon, A. Marandi, Y. Haribara, R. Hamerly, C. Langrock, S. Tamate, T. Inagaki, H. Takesue, S. Utsunomiya, K. Aihara, et al., Science 354, 614 (2016).
- [9]T. Inagaki, Y. Haribara, K. Igarashi, T. Sonobe, S. Tamate, T. Honjo, A. Marandi, P. L. McMahon, T. Umeki, K. E. Eubutsu, et al., Science 354, 603 (2016).
- [10]Y. Yamamoto, K. Aihara, T. Leleu, K.-i. Kawarabayashi, S. Kako, M. Fejer, K. Inoue, and H. Takesue, npj Quantum Information 3, 1 (2017).
- [11]hara, et al., Science advances 7, eabh0952 (2021).
- [12]F. Cai, S. Kumar, T. Van Vaerenbergh, X. Sheng, R. Liu, C. Li, Z. Liu, M. Foltin, S. Yu, Q. Xia, et al., Nature Electronics 3, 409 (2020).
- [13]M. Babaeian, D. T. Nguyen, V. Demir, M. Akbulut, P.-A. Blanche, Y. Kaneda, S. Guha, M. A. Neifeld, and N. Peyghambarian, Nature communications 10, 1 (2019).
- [14]V. Pal, S. Mahler, C. Tradonsky, A. A. Friesen, and N. Davidson, Physical Review Research 2, 033008 (2020).
- [15]M. Parto, W. Hayenga, A. Marandi, D. N. Christodoulides, and M. Khajavikhan, Nature materials 19, 725 (2020).
- [16]D. Pierangeli, G. Marcucci, and C. Conti, Physical review letters 122, 213902 (2019).
- [17]C. Roques-Carmes, Y. Shen, C. Zanoci, M. Prabhu, F. Atieh, J. Ding, T. Dubček, C. Mao, M. R. Johnson, V. Čeperić, et al., Nature communications 11, 249 (2020).
- [18]M. Vretenar, B. Kassenberg, S. Bissesar, C. Toebes, and J. Klaers, Physical Review Research 3, 023167 (2021).
- [19]K. Kalinin, G. Mourgias-Alexandris, H. Ballani, N. G. Berloff, J. H. Clegg, D. Cletheroe, C. Gkantsidis, I. Haller, V. Lyutsarev, F. Parmigiani, L. Pickup, et al., arXiv preprint arXiv:2304.12594 (2023).arxiv.org/abs/2304.12594
- [20]H. Goto, K. Endo, M. Suzuki, Y. Sakai, T. Kanao, Y. Hamakawa, R. Hidaka, M. Yamasaki, and K. Tatsumura, Science Advances 7, eabe7953 (2021).
- [21]P. Date, D. Arthur, and L. Pusey-Nazzaro, Scientific reports 11, 10029 (2021).
- [22]M. Gilli, D. Maringer, and E. Schumann, Numerical methods and optimization in finance (Academic Press, 2019).
- [23]N. A. Pierce and E. Winfree, Protein engineering 15, 779 (2002).
- [24]K. A. Dill, S. B. Ozkan, M. S. Shell, and T. R. Weikl, Annu. Rev. Biophys. 37, 289 (2008).
- [25]M. Calvanese Strinati and C. Conti, Nature Communications 13, 7248 (2022).
- [26]M. C. Strinati and C. Conti, Physical Review Letters 132, 017301 (2024).
- [27]K. P. Kalinin and N. G. Berloff, New Journal of Physics 20, 113023 (2018).
- [28]N. Stroev and N. G. Berloff, Physical Review Letters 126, 050504 (2021).
- [29]I. G. Rosenberg, (1975).
- [30]E. Boros and P. L. Hammer, Discrete applied mathematics 123, 155 (2002).
- [31]J. J. Hopfield, Proceedings of the national academy of sciences 79, 2554 (1982).
- [32]J. J. Hopfield and D. W. Tank, Biological cybernetics 52, 141 (1985).
- [33]H. Goto, Scientific reports 6, 1 (2016).
- [34]J. S. Cummins, H. Salman, and N. G. Berloff, arXiv preprint arXiv:2311.17359 (2023).DOI
- [35]K. P. Kalinin and N. G. Berloff, Communications Physics 5, 20 (2022).
- [36]A. Orvieto and A. Lucchi, Advances in Neural Information Processing Systems 32 (2019).
- [37]S. Saab Jr, S. Phoha, M. Zhu, and A. Ray, Machine Learning 111, 3245 (2022).
- [38]D. Subires, F. J. Gómez-Ruiz, A. Ruiz-García, D. Alonso, and A. Del Campo, Physical Review Research 4, 023104 (2022)
- [39]J.-R. Jiang and C.-W. Chu, in 2022 IEEE 4th ECICE (IEEE, 2022) pp. 406–411.
- [40]The BFGS algorithm is implemented with minimizer FindMinimum in Wolfram Mathematica. VISA Hamiltonian H_VISA is subject to BFGS with fixed γ_i = γ = 1.0Vi and P = 0.8.
- [41]J. Gancio and N. Rubido, Chaos, Solitons & Fractals 158, 112001 (2022).
- [42]Y. Haribara, H. Ishikawa, S. Utsunomiya, K. Aihara, and Y. Yamamoto, Quantum Science and Technology 2, 044002 (2017).
- [43]R. Hamerly, T. Inagaki, P. L. McMahon, D. Venturelli, A. Marandi, T. Onodera, E. Ng, C. Langrock, K. Inaba, T. Honjo, et al., Science advances 5, eaau0823 (2019).
- [44]M. P. Harrigan, K. J. Sung, M. Neeley, K. J. Satzinger, F. Arute, K. Arya, J. Atalaya, J. C. Bardin, R. Barends, S. Boixo, et al., Nature Physics 17, 332 (2021).
- [45]F. Böhm, G. Verschaffelt, and G. Van der Sande, Nature communications 10, 1 (2019).
- [46]S. Arora, E. Berger, H. Elad, G. Kindler, and M. Safra, in 46th Annual IEEE FOCS Symposium (IEEE, 2005) pp. 206–215.
- [47]Gurobi Optimization, LLC, “Gurobi Optimizer Reference Manual,” (2023).