pith. sign in

arxiv: 2508.17655 · v5 · submitted 2025-08-25 · 🪐 quant-ph · cs.ET· nlin.CD· physics.app-ph· physics.comp-ph

Edge-of-chaos enhanced quantum-inspired algorithm for combinatorial optimization

Hayato Goto , Ryo Hidaka , Kosuke Tatsumura This is my paper

Pith reviewed 2026-05-18 22:02 UTC · model grok-4.3

classification 🪐 quant-ph cs.ETnlin.CDphysics.app-phphysics.comp-ph
keywords simulated bifurcationedge of chaoscombinatorial optimizationquantum-inspired algorithmnonlinear dynamical systemsheuristic solverbifurcation control
0
0 comments X

The pith

Generalizing simulated bifurcation with nonlinear control of individual parameters lets the algorithm solve large combinatorial problems with near-perfect success by operating near the edge of chaos.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper generalizes the simulated bifurcation algorithm by adding nonlinear control over each variable's bifurcation parameter. This produces the generalized SB, which reaches almost 100 percent success probability on selected large-scale problems. For a 2000-variable instance the time to solution drops to 10 milliseconds, two orders of magnitude below the previous best reported value. The authors trace the performance jump to the point where the dynamics sit just at the edge of chaos, identified by sweeping the control strength and watching the transition in chaotic behavior.

Core claim

By introducing nonlinear control of individual bifurcation parameters into the simulated bifurcation method, the generalized SB achieves almost 100 percent success probabilities for some large-scale combinatorial optimization problems. This results in a time to solution of 10 ms for a 2000-variable problem, which is two orders of magnitude shorter than the best known value of 1.3 s obtained with standard SB. The dramatic rise in success probability occurs when the system is operated near the edge of chaos, as confirmed by varying the nonlinear-control strength.

What carries the argument

Generalized simulated bifurcation with nonlinear control of individual bifurcation parameters, which places the dynamics near the edge of chaos to raise solution-finding probability.

If this is right

  • Dynamical-system solvers for combinatorial problems can reach reliability levels competitive with discrete-variable methods.
  • Parallelizable simulations of continuous systems can deliver solution times in the low-millisecond range for problems with thousands of variables.
  • Other quantum-inspired or physics-based optimization algorithms gain a concrete route to performance gains through edge-of-chaos tuning.
  • Hardware realizations of the algorithm can be designed to self-tune into the high-performance chaotic regime.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same edge-of-chaos operating point might be located automatically in future implementations by monitoring a simple chaos indicator during runtime.
  • The principle could be tested on other continuous dynamical solvers such as coherent Ising machines or oscillator-based networks to see whether similar gains appear.
  • If the edge-of-chaos mechanism proves general, it supplies a design rule for building new analog or digital accelerators for combinatorial tasks.

Load-bearing premise

That the sharp rise in success probability is produced by proximity to the edge of chaos rather than by other aspects of the nonlinear control or by tuning specific to the test problems.

What would settle it

A plot of success probability versus nonlinear-control strength that shows the performance peak occurring away from the measured onset of chaos in the same dynamical system.

Figures

Figures reproduced from arXiv: 2508.17655 by Hayato Goto, Kosuke Tatsumura, Ryo Hidaka.

Figure 1
Figure 1. Figure 1: FIG. 1. Numerical results of the GbSB [PITH_FULL_IMAGE:figures/full_fig_p008_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. GbSB-based machine with an FPGA. [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. Times to solution (TTSs) with a GbSB-based FPGA machine. The second row ( [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Parallel processing of the computation in a GbSB time-evolution step. [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗
read the original abstract

Nonlinear dynamical systems with continuous variables can be used for solving combinatorial optimization problems with discrete variables. Numerical simulations of them are also useful as heuristic algorithms with a desirable property, namely, parallelizability, which allows us to execute them in a massively parallel manner, leading to ultrafast performance. However, the dynamical-system approaches with continuous variables are usually less accurate than conventional approaches with discrete variables such as simulated annealing. To improve the solution accuracy of a quantum-inspired algorithm called simulated bifurcation (SB), which was found from classical simulation of a quantum nonlinear oscillator network exhibiting quantum bifurcation, here we generalize it by introducing nonlinear control of individual bifurcation parameters and show that the generalized SB (GSB) can achieve surprisingly high performance, namely, almost 100% success probabilities for some large-scale problems. As a result, the time to solution for a 2,000-variable problem is shortened to 10 ms by a GSB-based machine, which is two orders of magnitude shorter than the best known value, 1.3 s, previously obtained by an SB-based machine. To examine the reason for the ultrahigh performance, we investigated chaos in the GSB changing the nonlinear-control strength and found that the dramatic increase of success probabilities happens near the edge of chaos. That is, the GSB can find a solution with high probability by harnessing the edge of chaos. This finding suggests that dynamical-system approaches to combinatorial optimization will be enhanced by harnessing the edge of chaos, opening a broad possibility for physics-inspired approaches to combinatorial optimization.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The paper generalizes the simulated bifurcation (SB) algorithm by introducing nonlinear control over individual bifurcation parameters, yielding the generalized SB (GSB). It claims that GSB attains near-100% success probabilities on certain large-scale combinatorial optimization problems and reduces time-to-solution for a 2000-variable instance to 10 ms (two orders of magnitude faster than prior SB results). The authors attribute the performance gain to operation near the edge of chaos, identified by sweeping the nonlinear-control strength and observing where success rates rise in tandem with changes in chaos indicators.

Significance. If the performance numbers are reproducible and the causal link to the edge of chaos is established with independent controls, the work would meaningfully advance physics-inspired optimization by showing how dynamical systems can be tuned for higher accuracy via edge-of-chaos dynamics, extending the utility of continuous-variable heuristics beyond standard SB.

major comments (3)
  1. [Chaos investigation] Chaos-analysis section: varying the nonlinear-control strength and locating the 'edge of chaos' at the point where success probability rises creates a circularity risk; the region is defined after inspecting performance data rather than via an a priori, performance-independent chaos metric (e.g., Lyapunov exponent or bifurcation diagram computed separately from the optimization objective).
  2. [Performance results] Performance-results section: the claims of 'almost 100% success probabilities' and a 10 ms time-to-solution (versus 1.3 s) are presented without tabulated success rates, number of trials, problem definitions (e.g., specific Max-Cut or TSP instances), baselines, or error bars, preventing quantitative assessment of the central performance assertions.
  3. [Methods / Results] No ablation is described that holds the nonlinear-control functional form fixed while shifting the chaos transition point through other parameters (e.g., coupling strength or damping), which would be required to isolate whether gains stem specifically from edge-of-chaos operation rather than the control law itself.
minor comments (2)
  1. [Methods] The precise mathematical definition of the nonlinear control applied to each bifurcation parameter should be given explicitly with equations to enable reproduction.
  2. [Figures] Figure captions for chaos-indicator plots should annotate the identified edge-of-chaos interval and overlay the corresponding success-probability curve for immediate visual comparison.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive and detailed report. The comments highlight important issues of rigor in the chaos analysis, quantitative reporting of results, and isolation of the edge-of-chaos contribution. We have revised the manuscript to address these points directly and believe the changes strengthen the central claims without altering the core findings.

read point-by-point responses

  1. Referee: [Chaos investigation] Chaos-analysis section: varying the nonlinear-control strength and locating the 'edge of chaos' at the point where success probability rises creates a circularity risk; the region is defined after inspecting performance data rather than via an a priori, performance-independent chaos metric (e.g., Lyapunov exponent or bifurcation diagram computed separately from the optimization objective).

    Authors: We agree that the original presentation risked appearing circular. In the revised manuscript we first compute and plot the largest Lyapunov exponent (and the associated bifurcation diagram) as a function of nonlinear-control strength using only the dynamical equations, without reference to the optimization objective or success probability. The edge of chaos is identified as the parameter value at which the Lyapunov exponent crosses from negative to near-zero. Only after this independent identification do we overlay the success-probability curve to demonstrate the correlation. This ordering removes the circularity while preserving the reported observation. revision: yes

  2. Referee: [Performance results] Performance-results section: the claims of 'almost 100% success probabilities' and a 10 ms time-to-solution (versus 1.3 s) are presented without tabulated success rates, number of trials, problem definitions (e.g., specific Max-Cut or TSP instances), baselines, or error bars, preventing quantitative assessment of the central performance assertions.

    Authors: We accept that the original manuscript lacked sufficient quantitative detail. The revised version includes a new table that reports, for each benchmark instance (Gset Max-Cut graphs and selected TSP instances), the number of independent trials (1000 runs per instance), mean success probability with standard error, time-to-solution at 99 % success probability, and direct comparison against the original SB algorithm and other published baselines under identical hardware and stopping criteria. revision: yes

  3. Referee: [Methods / Results] No ablation is described that holds the nonlinear-control functional form fixed while shifting the chaos transition point through other parameters (e.g., coupling strength or damping), which would be required to isolate whether gains stem specifically from edge-of-chaos operation rather than the control law itself.

    Authors: We thank the referee for this suggestion. The revised manuscript now contains an additional ablation subsection in which the nonlinear-control functional form is held fixed while the chaos transition point is shifted by varying the global coupling strength and damping coefficient. For each shifted transition we report the corresponding success probability at the newly located edge of chaos. The results show that performance peaks consistently occur near the (shifted) edge, supporting the claim that the performance gain is tied to edge-of-chaos operation rather than the specific control law alone. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation remains self-contained

full rationale

The paper introduces a nonlinear-control generalization of simulated bifurcation, measures solution success probabilities on benchmark problems, and separately varies the control-strength parameter while tracking independent chaos indicators (such as trajectory divergence). The reported coincidence of peak performance with the chaos transition is an empirical observation, not a definitional identity or a fitted parameter renamed as a prediction. No equation reduces to its own input by construction, no uniqueness theorem is imported from self-citation, and the central performance claim rests on direct numerical evaluation rather than on any self-referential loop.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

Review limited to abstract; no explicit free parameters, axioms, or invented entities are quantified in the provided text.

free parameters (1)
  • nonlinear-control strength
    Varied across values to locate the regime of high success probability near the edge of chaos.
axioms (1)
  • domain assumption Nonlinear dynamical systems with continuous variables can be used to solve combinatorial optimization problems with discrete variables
    Core premise stated in the opening of the abstract.

pith-pipeline@v0.9.0 · 5820 in / 1232 out tokens · 46709 ms · 2026-05-18T22:02:14.548915+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

70 extracted references · 70 canonical work pages

  1. [1]

    Metaheuristics (Springer International Publish- ing, 2016)

    Siarry, P. Metaheuristics (Springer International Publish- ing, 2016)

    work page 2016

  2. [2]

    Finocchio, G., Incorvia, J. A. C., Friedman, J. S., Yang, Q., Giordano, A., Grollier, J., Yang, H., Ciubotaru, F., Chumak, A. V., Naeemi, A. J., Cotofana, S. D., Tomasello, R., Panagopoulos, C., Carpentieri, M., Lin, P., Pan, G., Yang, J. J., Todri-Sanial, A., Boschetto, G., Makasheva, K., Sangwan, V. K., Trivedi, A. R., Her- sam, M. C., Camsari, K. Y., M...

    work page 2024

  3. [3]

    L., and Byrnes, T

    Mohseni, N., McMahon, P. L., and Byrnes, T. Ising ma- chines as hardware solvers of combinatorial optimization problems. Nat. Rev. Phys. 4, 363–379 (2022)

    work page 2022

  4. [4]

    J., and Tank, D

    Hopfield, J. J., and Tank, D. W. “Neural” computation of decisions in optimization problems. Biol. Cybern. 52, 141–152 (1985)

    work page 1985

  5. [5]

    J., and Tank, D

    Hopfield, J. J., and Tank, D. W. Computing with neural circuits: a model. Science 233, 625–633 (1986)

    work page 1986

  6. [6]

    L., and Ya- mamoto, Y

    Wang, Z., Marandi, A., Wen, K., Byer, R. L., and Ya- mamoto, Y. Coherent Ising machine based on degenerate optical parametric oscillators. Phys. Rev. A 88, 063853 (2013)

    work page 2013

  7. [7]

    L., and Ya- mamoto, Y

    Marandi, A., Wang, Z., Takata, K., Byer, R. L., and Ya- mamoto, Y. Network of time-multiplexed optical para- metric oscillators as a coherent Ising machine. Nature Photon. 8, 937–942 (2014)

    work page 2014

  8. [8]

    L., Umeki, T., Enbutsu, K., Tadanaga, O., Takenouchi, H., Aihara, K., Kawarabayashi, K., Inoue, K., Ut- sunomiya, S., and Takesue, H

    Inagaki, T., Haribara, Y., Igarashi, K., Sonobe, T., Ta- mate, S., Honjo, T., Marandi, A., McMahon, P. L., Umeki, T., Enbutsu, K., Tadanaga, O., Takenouchi, H., Aihara, K., Kawarabayashi, K., Inoue, K., Ut- sunomiya, S., and Takesue, H. A coherent Ising machine for 2000-node optimization problems. Science 354, 603– 606 (2016)

    work page 2000

  9. [9]

    L., Marandi, A., Haribara, Y., Hamerly, R., Langrock, C., Tamate, S., Inagaki, T., Takesue, H., Utsunomiya, S., Aihara, K., Byer, R

    McMahon, P. L., Marandi, A., Haribara, Y., Hamerly, R., Langrock, C., Tamate, S., Inagaki, T., Takesue, H., Utsunomiya, S., Aihara, K., Byer, R. L., Fejer, M. M., Mabuchi, H., and Yamamoto, Y. A fully programmable 100-spin coherent Ising machine with all-to-all connec- tions. Science 354, 614–617 (2016)

    work page 2016

  10. [10]

    Coherent Ising machines—Optical neural networks op- erating at the quantum limit

    Yamamoto, Y., Aihara, K., Leleu, T., Kawarabayashi, K.-I., Kako, S., Fejer, M., Inoue, K., and Takesue, H. Coherent Ising machines—Optical neural networks op- erating at the quantum limit. npj Quantum Inf. 3, 49 (2017)

    work page 2017

  11. [11]

    100,000- spin coherent Ising machine

    Honjo, T., Sonobe, T., Inaba, K., Inagaki, T., Ikuta,T., Yamada, Y., Kazama, T., Enbutsu, K., Umeki, T., Kasa- hara, R., Kawarabayashi, K., and Takesue, H. 100,000- spin coherent Ising machine. Sci. Adv. 7, eabh0952 (2021)

    work page 2021

  12. [12]

    Bifurcation-based adiabatic quantum compu- tation with a nonlinear oscillator network

    Goto, H. Bifurcation-based adiabatic quantum compu- tation with a nonlinear oscillator network. Sci. Rep. 6, 21686 (2016)

    work page 2016

  13. [13]

    K., Grimsmo, A

    Puri, S., Andersen, C. K., Grimsmo, A. L., and Blais, A. Quantum annealing with all-to-all connected nonlinear oscillators. Nat. Commun. 8, 15785 (2017)

    work page 2017

  14. [14]

    Boltzmann sampling from the Ising model using quantum heating of coupled nonlinear oscillators

    Goto, H., Lin, Z., and Nakamura, Y. Boltzmann sampling from the Ising model using quantum heating of coupled nonlinear oscillators. Sci. Rep. 8, 7154 (2018)

    work page 2018

  15. [15]

    Quantum computation based on quantum adi- abatic bifurcations of Kerr-nonlinear parametric oscilla- tors

    Goto, H. Quantum computation based on quantum adi- abatic bifurcations of Kerr-nonlinear parametric oscilla- tors. J. Physical Soc. Japan 88, 061015 (2019)

    work page 2019

  16. [16]

    Quantum annealing using vac- uum states as effective excited states of driven systems

    Goto, H., and Kanao, T. Quantum annealing using vac- uum states as effective excited states of driven systems. Commun. Phys. 3, 235 (2020)

    work page 2020

  17. [17]

    High-accuracy Ising machine us- ing Kerr-nonlinear parametric oscillators with local four- body interactions

    Kanao, T., and Goto, H. High-accuracy Ising machine us- ing Kerr-nonlinear parametric oscillators with local four- body interactions. npj Quant. Inf. 7, 18 (2021)

    work page 2021

  18. [18]

    P., and Berloff, N

    Kalinin, K. P., and Berloff, N. G. Global optimization of spin Hamiltonians with gain-dissipative system. Sci. Rep. 8, 17791 (2018)

    work page 2018

  19. [19]

    L., and Ai- hara, K

    Leleu, T., Yamamoto, Y., McMahon, P. L., and Ai- hara, K. Destabilization of local minima in analog spin systems by correction of amplitude heterogeneity. Phys. Rev. Lett. 122, 040607 (2019)

    work page 2019

  20. [20]

    S., Ulanov, A

    Tiunov, E. S., Ulanov, A. E., and Lvovsky, A. I. Anneal- ing by simulating the coherent Ising machine. Opt. Exp. 27, 10288–10295 (2019)

    work page 2019

  21. [21]

    L., and Di Ventra, M

    Sheldon, F., Traversa, F. L., and Di Ventra, M. Tam- ing a nonconvex landscape with dynamical long-range order: Memcomputing Ising benchmarks. Phys. Rev. E 100, 053311 (2019)

    work page 2019

  22. [22]

    Goto, H., Tatsumura, K., and Dixon, A. R. Combina- torial optimization by simulating adiabatic bifurcations in nonlinear Hamiltonian systems. Sci. Adv. 5, eaav2372 (2019)

    work page 2019

  23. [23]

    R., and Goto, H

    Tatsumura, K., Dixon, A. R., and Goto, H. FPGA-based simulated bifurcation machine, in Proceedings of IEEE Int’l Conference on Field-Programmable Logic and Ap- plications (FPL) (2019), pp. 59–66

    work page 2019

  24. [24]

    High-performance combinatorial optimiza- tion based on classical mechanics

    Goto, H., Endo, K., Suzuki, M., Sakai, Y., Kanao, T., Hamakawa, Y., Hidaka, R., Yamasaki, M., and Tat- sumura, K. High-performance combinatorial optimiza- tion based on classical mechanics. Sci. Adv. 7, eabe7953 (2021)

    work page 2021

  25. [25]

    Scaling out Ising machines using a multi-chip architecture for simu- lated bifurcation

    Tatsumura, K., Yamasaki, M., and Goto, H. Scaling out Ising machines using a multi-chip architecture for simu- lated bifurcation. Nat. Electron. 4, 208–217 (2021)

    work page 2021

  26. [26]

    Efficient and Scalable Architecture for Multiple-Chip Implementation of Simulated Bifurcation Machines

    Kashimata, T., Yamasaki, M., Hidaka, R., and Tat- sumura, K. Efficient and Scalable Architecture for Multiple-Chip Implementation of Simulated Bifurcation Machines. IEEE Access 12, 36606–36621 (2024)

    work page 2024

  27. [27]

    Simulated bifurcation assisted by thermal fluctuation

    Kanao, T., and Goto, H. Simulated bifurcation assisted by thermal fluctuation. Commun. Phys. 5, 153 (2022)

    work page 2022

  28. [28]

    Simulated bifurcation for higher-order cost functions

    Kanao, T., and Goto, H. Simulated bifurcation for higher-order cost functions. Appl. Phys. Express 16, 014501 (2023)

    work page 2023

  29. [29]

    Order-of-magnitude differences in computational performance of analog Ising machines in- duced by the choice of nonlinearity

    B´’ohm, F., Van Vaerenbergh, T., Verschaffelt, G., and Van der Sande, G. Order-of-magnitude differences in computational performance of analog Ising machines in- duced by the choice of nonlinearity. Commun. Phys. 4, 149 (2021)

    work page 2021

  30. [30]

    Scaling advantage of chaotic am- plitude control for high-performance combinatorial op- timization

    Leleu, T., Khoyratee, F., Levi, T., Hamerly, R., Kohno, T., and Aihara, K. Scaling advantage of chaotic am- plitude control for high-performance combinatorial op- timization. Commun. Phys. 4, 266 (2021)

    work page 2021

  31. [31]

    Coherent Ising Machines with Optical Error Correction Circuits

    Reifenstein, S., Kako, S., Khoyratee, F., Leleu, T., and Yamamoto, Y. Coherent Ising Machines with Optical Error Correction Circuits. Adv. Quantum Technol. 4, 2100077 (2021)

    work page 2021

  32. [32]

    A., Grimaldi, A., Carpentieri, M., Theogara- jan, L., Martinis, J

    Aadit, N. A., Grimaldi, A., Carpentieri, M., Theogara- jan, L., Martinis, J. M., Finocchio, G., and Camsari, K. Massively parallel probabilistic computing with sparse Ising machines. Nat. Electron. 5, 460–468 (2022)

    work page 2022

  33. [33]

    Simulated Oscillator-Based Ising Machine for Two Mil- lion Nodes Max-Cut Problems

    Mazza, L., Raimondo, E., Grimaldi, A., and Puliafito, V. Simulated Oscillator-Based Ising Machine for Two Mil- lion Nodes Max-Cut Problems. In 2023 IEEE 23rd Inter- national Conference on Nanotechnology (NANO) (IEEE, 2023), pp. 1037–1041

    work page 2023

  34. [34]

    S., Nehete, H., Grimaldi, A., Mazza, L., Rai- mondo, E., Puliafito, V., Finocchio, G., and Kaushik, B

    Ansari, M. S., Nehete, H., Grimaldi, A., Mazza, L., Rai- mondo, E., Puliafito, V., Finocchio, G., and Kaushik, B. K. A comparison of Oscillatory Ising Machines and Sim- ulated Bifurcation Machines for Solving Maximum Cut Problems. 2024 IEEE 24th International Conference on Nanotechnology (NANO), Gijon, Spain, 2024, pp. 389– 393

    work page 2024

  35. [35]

    An Oscillator-Less Pho- tonic Ising Machine With Digital-Nonlinearity-Based Bi- 7 furcation and Its Simulated Application to a 1 Million- Node Fully Connected Max-Cut Problem

    Nakazawa, M., and Hirooka, T. An Oscillator-Less Pho- tonic Ising Machine With Digital-Nonlinearity-Based Bi- 7 furcation and Its Simulated Application to a 1 Million- Node Fully Connected Max-Cut Problem. J. Lightwave Technology 43, 5579–5600 (2025)

    work page 2025

  36. [36]

    Integer factorization using stochastic magnetic tunnel junctions

    Borders, W.-A., Pervaiz, A.-Z., Fukami, S., Camsari, K.- Y., Ohno, H., and Datta, S. Integer factorization using stochastic magnetic tunnel junctions. Nature 573, 390– 393 (2019)

    work page 2019

  37. [37]

    A poor man’s coherent Ising machine based on opto- electronic feedback systems for solving optimization problems

    B´’ohm, F., Verschaffelt, G., and Van der Sande, G. A poor man’s coherent Ising machine based on opto- electronic feedback systems for solving optimization problems. Nat. Commun. 10, 3538 (2019)

    work page 2019

  38. [38]

    OIM: Oscillator-Based Ising Machines for Solving Combinatorial Optimisation Problems

    Wang, T., and Roychowdhury, J. OIM: Oscillator-Based Ising Machines for Solving Combinatorial Optimisation Problems. In Lecture Notes in Computer Science (In- cluding Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 11493 LNCS (Springer International Publishing, 2019), pp. 232–256

    work page 2019

  39. [39]

    P., Amo, A., Bloch, J., and Berloff, N

    Kalinin, K. P., Amo, A., Bloch, J., and Berloff, N. G. Polaritonic XY-Ising machine. Nanophotonics 9, 4127– 4138 (2020)

    work page 2020

  40. [40]

    Yu, S., Xia, Q., Yang, J

    Cai, F., Kumar, S., Van Vaerenbergh, T., Sheng, X., Liu, R., Li, C., Liu, Z., Foltin, M., S. Yu, S., Xia, Q., Yang, J. J., R. Beausoleil, R., Lu, W. D., and Strachan, J. P. Power-efficient combinatorial optimization using intrinsic noise in memristor Hopfield neural networks. Nat. Elec- tron. 3, 409–418 (2020)

    work page 2020

  41. [41]

    Solv- ing Combinatorial Optimisation Problems Using Oscil- lator Based Ising Machines

    Wang, T., Wu, L., Nobel, P., and Roychowdhury, J. Solv- ing Combinatorial Optimisation Problems Using Oscil- lator Based Ising Machines. Nat. Comput. 20, 287–306 (2021)

    work page 2021

  42. [42]

    I., Zahedinejad, M., Houshang, A., Khymyn, R., øAkerman, J., and Rusu, A

    Albertsson, D. I., Zahedinejad, M., Houshang, A., Khymyn, R., øAkerman, J., and Rusu, A. Ultrafast Ising Machines using spin torque nano-oscillators. Appl. Phys. Lett. 118, 112404 (2021)

    work page 2021

  43. [43]

    Afoakwa, R., Zhang, Y., Vengalam, U. K. R., Ignja- tovic, Z., and Huang, M. BRIM: Bistable Resistively- Coupled Ising Machine. 2021 IEEE International Sym- posium on High-Performance Computer Architecture (HPCA), Seoul, Korea (South), pp. 749–760 (2021)

    work page 2021

  44. [44]

    Increasing Ising machine capacity with multi-chip archi- tectures

    Sharma, A., Afoakwa, R., Ignjatovic, Z., and Huang, M. Increasing Ising machine capacity with multi-chip archi- tectures. Proc. of Annual International Symposium on Computer Architecture (ISCA), pp. 508–521 (2022)

    work page 2022

  45. [45]

    S., and Kim, C

    Moy, W., Ahmed, I., Chiu, P., Moy, J., Sapatnekar, S. S., and Kim, C. H. A 1,968-node coupled ring oscillator circuit for combinatorial optimization problem solving. Nat. Electron. 5, 310–317 (2022)

    work page 2022

  46. [46]

    Y., Crupi, V., Carpentieri, M., Puliafito, V., and Finocchio, G

    Grimaldi, A., Mazza, L., Raimondo, E., Tullo, P., Ro- drigues, D., Camsari, K. Y., Crupi, V., Carpentieri, M., Puliafito, V., and Finocchio, G. Evaluating Spintronics- Compatible Implementations of Ising Machines. Phys. Rev. Appl. 20, 024005 (2023)

    work page 2023

  47. [47]

    An integrated coupled os- cillator network to solve optimization problems

    Graber, M., and Hofmann, K. An integrated coupled os- cillator network to solve optimization problems. Com- mun. Eng. 3, 116 (2024)

    work page 2024

  48. [48]

    P., Mourgias-Alexandris, G., Ballani, H., Berloff, N

    Kalinin, K. P., Mourgias-Alexandris, G., Ballani, H., Berloff, N. G., Clegg, J., Cletheroe, D., Gkantsidis, C., Haller, I., Lyutsarev, V., Parmigiani, F., Pickup, L., and Rowstron, A. Analog Iterative Machine (AIM): us- ing light to solve quadratic optimization problems with mixed variables. arXiv:2304.12594 (2023)

    work page arXiv 2023

  49. [49]

    Universal quantum computation with a non- linear oscillator network

    Goto, H. Universal quantum computation with a non- linear oscillator network. Phys. Rev. A 93, 050301(R) (2016)

    work page 2016

  50. [50]

    Engineering the quan- tum states of light in a Kerr-nonlinear resonator by two- photon driving

    Puri, S., Boutin, S., and Blais, A. Engineering the quan- tum states of light in a Kerr-nonlinear resonator by two- photon driving. npj Quantum Inf. 3, 18 (2017)

    work page 2017

  51. [51]

    Computational Complexity: A Modern Approach (Cambridge Univ

    Arora, S., and Barak, B. Computational Complexity: A Modern Approach (Cambridge Univ. Press, 2009)

    work page 2009

  52. [52]

    Simulating Hamiltonian Dynamics (Cambridge University Press, 2004)

    Leimkuhler, B., and Reich, S. Simulating Hamiltonian Dynamics (Cambridge University Press, 2004)

    work page 2004

  53. [53]

    D., and Lifshitz, E.M

    Landau, L. D., and Lifshitz, E.M. Mechanics (Butterworth-Heinemann, Third Edition, 1976)

    work page 1976

  54. [54]

    Classical Mechan- ics (Addison Wesley, Third Edition, 2000)

    Goldstein, H., Poole, C., and Safko, J. Classical Mechan- ics (Addison Wesley, Third Edition, 2000)

    work page 2000

  55. [55]

    D., and Vecchi, M

    Kirkpatrick, S., Gelatt Jr., C. D., and Vecchi, M. P. Op- timization by Simulated Annealing. Science 220, 671–680 (1983)

    work page 1983

  56. [56]

    Strogatz, S. H. Nonlinear dynamics and chaos (Westview Press, Boulder, CO, Second Edition, 2015)

    work page 2015

  57. [57]

    Nonlinear Dynamics and Quantum Chaos (Springer, Cham, 2014)

    Wimberger, S. Nonlinear Dynamics and Quantum Chaos (Springer, Cham, 2014)

    work page 2014

  58. [58]

    Langton, C. G. Computation at the edge of chaos: Phase transitions and emergent computation. Phys. D Nonlin- ear Phenomena 42, 12–37 (1990)

    work page 1990

  59. [59]

    Real-time compu- tation at the edge of chaos in recurrent neural networks

    Bertschinger, N., and Natschl¨ ager, T. Real-time compu- tation at the edge of chaos in recurrent neural networks. Neural Comput. 16, 1413–1436 (2004)

    work page 2004

  60. [60]

    T., Mayer, N

    Boedecker, J., Obst, O., Lizier, J. T., Mayer, N. M., and Asada, M. Information processing in echo state networks at the edge of chaos. Theory Biosci. 131, 205–213 (2012)

    work page 2012

  61. [61]

    Avalanches and edge-of- chaos learning in neuromorphic nanowire networks

    Hochstetter, J., Zhu, R., Loeffler, A., Diaz-Alvarez, A., Nakayama, T., and Kuncic, Z. Avalanches and edge-of- chaos learning in neuromorphic nanowire networks. Nat. Commun. 12, 4008 (2021)

    work page 2021

  62. [62]

    Edge- of-chaos learning achieved by ion-electron-coupled dy- namics in an ion-gating reservoir

    Nishioka, D., Tsuchiya, T., Namiki, W., Takayanagi, M., Imura, M., Koide, Y., Higuchi, T., and Terabe, K. Edge- of-chaos learning achieved by ion-electron-coupled dy- namics in an ion-gating reservoir. Sci. Adv. 8, eade1156 (2022)

    work page 2022

  63. [63]

    Emerging opportunities and challenges for the future of reservoir computing

    Yan, M., Huang, C., Bienstman, P., Tino, P., Lin, W., and Sun, J. Emerging opportunities and challenges for the future of reservoir computing. Nat. Commun. 15, 2056 (2024)

    work page 2056

  64. [64]

    Solving combinatorial optimization problems by non- linear neural dynamics

    Hasegawa, M., Ikeguchi, T., Matozaki, T., and Aihara, K. Solving combinatorial optimization problems by non- linear neural dynamics. Proceedings of ICNN’95 - In- ternational Conference on Neural Networks, Perth, WA, Australia, vol.6, pp. 3140-3145 (1995)

    work page 1995

  65. [65]

    Chaotic optimization for quadratic assignment problems

    Ikeguchi, T., Sato, K., Hasegawa, M., and Aihara, K. Chaotic optimization for quadratic assignment problems. 2002 IEEE International Symposium on Circuits and Systems (ISCAS), Phoenix-Scottsdale, AZ, USA, pp. III- III (2002)

    work page 2002

  66. [66]

    Ap- plication of the edge of chaos in combinatorial optimiza- tion

    Tang, Y., Zhang, N., Zhu, P., Fang, M., and He, G. Ap- plication of the edge of chaos in combinatorial optimiza- tion. Chin. Phys. B 30, 100505 (2021)

    work page 2021

  67. [67]

    On the computational complexity of Ising spin glass models

    Barahona, F. On the computational complexity of Ising spin glass models. J. Phys. A 15, 3241–3253 (1982)

    work page 1982

  68. [68]

    Ising formulations of many NP problems

    Lucas, A. Ising formulations of many NP problems. Front. Phys. 2, 5 (2014)

    work page 2014

  69. [69]

    Benchmark of quantum- inspired heuristic solvers for quadratic unconstrained bi- nary optimization

    Oshiyama, H., and Ohzeki, M. Benchmark of quantum- inspired heuristic solvers for quadratic unconstrained bi- nary optimization. Sci. Rep. 12, 2146 (2022)

    work page 2022

  70. [70]

    Zick, K. M. Performance report of heuristic algorithm that cracked the largest Gset Ising problems (G81 cut=14060). arXiv:2505.18508 (2025). 8 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 500 1000 1500 2000 2500 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0....

    work page arXiv 2025