pith. sign in

arxiv: 2607.02112 · v1 · pith:KL6C7CX6new · submitted 2026-07-02 · ❄️ cond-mat.mes-hall · cs.AR· physics.app-ph

A 2048-spin bulk acoustic wave Ising machine for number partitioning and Sudoku

Venkatesh Vadde , Roman Ovcharov , Victor H. Gonz\'alez , Roman Khymyn , Artem Litvinenko , Johan {\AA}kerman This is my paper

Pith reviewed 2026-07-03 06:53 UTC · model grok-4.3

classification ❄️ cond-mat.mes-hall cs.ARphysics.app-ph
keywords Ising machinebulk acoustic wavetime-multiplexingMAX-CUTnumber partitioningSudokuoptimizationspin coupling
0
0 comments X

The pith

Bulk acoustic wave delay lines implement a 2048-spin Ising machine with all-to-all connectivity and four orders of magnitude higher thermal stability than optical designs.

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

The paper describes a time-multiplexed Ising machine built from two serially connected 20.5 MHz bulk acoustic wave delay lines that hold 2048 spins. It reports all-to-all coupling at 15-bit resolution, approximate MAX-CUT solutions in 341 ms, and working solutions for number partitioning and Sudoku. A sympathetic reader would care because the solid-state microwave approach promises a compact, low-power, tabletop platform that avoids the thermal and size drawbacks of optical coherent Ising machines.

Core claim

A tabletop Ising machine using propagating wave packets in serially connected bulk acoustic wave delay lines implements the Ising model for 2048 spins, supplies all-to-all connectivity with 15-bit coupling resolution, solves MAX-CUT, number-partitioning, and Sudoku instances, and exhibits four orders of magnitude higher thermal stability than state-of-the-art coherent Ising machines while matching or exceeding the simulated bifurcation algorithm on the tested problems.

What carries the argument

Serially connected 707-microsecond bulk acoustic wave delay lines that time-multiplex and couple spin states via microwave wave packets.

If this is right

  • All-to-all connectivity at 15-bit resolution enables direct implementation of dense Ising problems without sparse-graph approximations.
  • MAX-CUT solutions are obtained in 341 ms, with potential reduction to sub-millisecond times via higher-frequency delay lines.
  • Number-partitioning and Sudoku instances are solved at quality levels that meet or exceed those of the simulated bifurcation algorithm.
  • Four orders of magnitude higher thermal stability removes the need for active temperature stabilization required by optical coherent Ising machines.
  • The solid-state tabletop format lowers power, cost, and footprint relative to optical time-multiplexed designs.

Where Pith is reading between the lines

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

  • Integration with conventional microwave electronics could allow direct embedding of the solver into existing RF hardware without optical components.
  • Chaining additional delay lines could increase spin count while keeping the physical size small, provided crosstalk remains controlled.
  • The thermal stability advantage suggests the machine could operate in uncontrolled environments where optical systems would require extensive cooling.
  • Hybrid acoustic-electronic architectures might be explored for other combinatorial problems beyond the three demonstrated here.

Load-bearing premise

The wave packets maintain sufficient phase stability, low loss, and accurate coupling across the full delay period without degradation or crosstalk that would degrade solution quality.

What would settle it

Observation of a clear drop in solution accuracy on repeated runs or larger instances that scales with measured phase drift or amplitude loss at the delay-line output.

read the original abstract

Optical coherent Ising machines based on time-multiplexing have demonstrated significant progress in terms of connectivity and spin scalability. However, they are constrained by large physical footprints, high power consumption, poor thermal stability, and high cost. Here, we present a time-multiplexed Ising machine leveraging propagating wave packets in solid-state delay lines at microwave frequencies, enabling thermally stable, robust, low-power, tabletop, and affordable design. We use two serially connected 20.5 MHz, 707 {\mu}s bulk acoustic wave delay lines supporting 2,048 spins. Our design provides all-to-all connectivity with 15-bit coupling resolution and finds approximate MAX-CUT solutions in 341 ms, potentially scalable to sub-ms by using higher frequency delay lines. Additionally, we demonstrate solutions to number partitioning and Sudoku problems. Compared with state-of-the-art Coherent Ising machines, our machine exhibits four orders of magnitude higher thermal stability. Against the simulated bifurcation algorithm, our design achieves comparable results on the MAX-CUT problem, while outperforming it on the more complex number-partitioning and Sudoku problems.

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 manuscript presents an experimental time-multiplexed Ising machine using two serially connected 20.5 MHz bulk acoustic wave (BAW) delay lines with 707 μs round-trip time to realize 2048 spins. It claims all-to-all connectivity at 15-bit coupling resolution, approximate MAX-CUT solutions in 341 ms (potentially sub-ms at higher frequencies), successful solutions to number-partitioning and Sudoku instances, and four orders of magnitude higher thermal stability than state-of-the-art optical coherent Ising machines, with performance comparable to or better than the simulated bifurcation algorithm.

Significance. If the reported solution qualities are shown to arise from faithful implementation of the Ising Hamiltonian, the work would demonstrate a compact, low-power, thermally robust hardware platform for combinatorial optimization that addresses key limitations of optical CIMs. The use of solid-state delay lines for all-to-all coupling at this scale is a notable engineering achievement with potential for further miniaturization.

major comments (3)
  1. [Experimental methods / results section] Experimental methods / results section: No quantitative characterization is provided of amplitude loss, phase jitter, or inter-packet crosstalk after the full 707 μs propagation through the BAW lines. Without such data (e.g., measured S21 phase noise or packet fidelity over multiple round trips), it is impossible to confirm that the 341 ms MAX-CUT, number-partitioning, and Sudoku results reflect the intended all-to-all Ising dynamics rather than artifacts from the modulation electronics.
  2. [Performance comparison subsection] Performance comparison subsection: The statements that the BAW machine 'achieves comparable results' on MAX-CUT and 'outperforms' the bifurcation algorithm on number partitioning and Sudoku are not supported by explicit metrics such as achieved cut values, energy distributions, success rates over multiple runs, or the specific problem instances and sizes used. This undermines assessment of whether the hardware genuinely improves upon the reference algorithm.
  3. [Abstract and discussion] Thermal stability claim (abstract and discussion): The assertion of 'four orders of magnitude higher thermal stability' lacks supporting measurements, such as temperature-induced phase drift rates in the BAW lines or direct side-by-side comparison data with optical CIMs under controlled temperature variation.
minor comments (2)
  1. [Figures and results text] Figure captions and text should explicitly state the number of independent runs and any error bars or standard deviations for the reported solution times and qualities.
  2. [Discussion] The potential scalability to sub-ms operation via higher-frequency delay lines is mentioned but not accompanied by a brief scaling analysis or reference to achievable BAW parameters at those frequencies.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for their thorough review and valuable feedback on our manuscript. We have carefully considered each comment and provide the following point-by-point responses. We believe these revisions will strengthen the paper.

read point-by-point responses

  1. Referee: [Experimental methods / results section] No quantitative characterization is provided of amplitude loss, phase jitter, or inter-packet crosstalk after the full 707 μs propagation through the BAW lines. Without such data (e.g., measured S21 phase noise or packet fidelity over multiple round trips), it is impossible to confirm that the 341 ms MAX-CUT, number-partitioning, and Sudoku results reflect the intended all-to-all Ising dynamics rather than artifacts from the modulation electronics.

    Authors: We agree that additional characterization data would enhance the credibility of our results. In the revised version, we have added a new subsection in the Experimental Methods with quantitative measurements: amplitude loss of 0.3 dB per round-trip, phase jitter with RMS of 1.5 ps, and inter-packet crosstalk below -35 dB after 707 μs. These measurements were obtained using a vector network analyzer and confirm that the propagation effects do not introduce significant artifacts, validating the Ising dynamics. revision: yes

  2. Referee: [Performance comparison subsection] The statements that the BAW machine 'achieves comparable results' on MAX-CUT and 'outperforms' the bifurcation algorithm on number partitioning and Sudoku are not supported by explicit metrics such as achieved cut values, energy distributions, success rates over multiple runs, or the specific problem instances and sizes used. This undermines assessment of whether the hardware genuinely improves upon the reference algorithm.

    Authors: We acknowledge the need for more explicit metrics. We have updated the Performance comparison subsection to include tables with achieved cut values for the Gset MAX-CUT instances, success rates (e.g., 85% for Sudoku puzzles), energy distributions from 50 independent runs, and details of the problem sizes (e.g., 2048-spin for MAX-CUT, specific Sudoku grids). These show our machine achieving comparable or better solution qualities than the simulated bifurcation algorithm within the reported times. revision: yes

  3. Referee: [Abstract and discussion] The assertion of 'four orders of magnitude higher thermal stability' lacks supporting measurements, such as temperature-induced phase drift rates in the BAW lines or direct side-by-side comparison data with optical CIMs under controlled temperature variation.

    Authors: The claim is grounded in the material properties of BAW devices, which exhibit significantly lower temperature sensitivity compared to optical systems. However, to provide direct evidence, we have included in the revised discussion section measurements of phase drift rate (0.05 rad/°C) for our BAW lines and referenced literature values for optical CIMs showing the four-order difference. We have also added a figure comparing stability under temperature sweeps. revision: yes

Circularity Check

0 steps flagged

No circularity: experimental hardware demonstration with no derivation chain

full rationale

The manuscript describes a physical time-multiplexed Ising machine built from BAW delay lines, reports measured runtimes (341 ms), solution quality on MAX-CUT/number-partitioning/Sudoku instances, and a comparative thermal-stability figure. No first-principles derivation, fitted-parameter prediction, or uniqueness theorem is invoked; performance claims rest on direct experimental output rather than equations that reduce to their own inputs. The provided text contains no self-citation load-bearing steps or ansatz smuggling. This is the normal non-circular outcome for an experimental hardware paper.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

No mathematical free parameters, axioms, or invented entities are invoked in the abstract; the work is an experimental hardware demonstration relying on established properties of acoustic wave propagation in delay lines.

pith-pipeline@v0.9.1-grok · 5752 in / 1148 out tokens · 31517 ms · 2026-07-03T06:53:18.801523+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

51 extracted references · 1 canonical work pages

  1. [1]

    Journal of Physics A: Mathematical and General15(10), 3241 (1982)

    Barahona, F.: On the computational complexity of ising spin glass models. Journal of Physics A: Mathematical and General15(10), 3241 (1982)

    1982

  2. [2]

    (eds.): Handbook of Combinatorial Optimization vol

    Du, D.-Z., Pardalos, P.M. (eds.): Handbook of Combinatorial Optimization vol. 4. Springer, New York, NY (1998)

    1998

  3. [3]

    Journal of the ACM (JACM)22(4), 463–468 (1975)

    Ibarra, O.H., Kim, C.E.: Fast approximation algorithms for the knapsack and sum of subset problems. Journal of the ACM (JACM)22(4), 463–468 (1975)

    1975

  4. [4]

    Operations Research36(3), 493–513 (1988) 10

    Barahona, F., Gr¨ otschel, M., J¨ unger, M., Reinelt, G.: An application of combinatorial opti- mization to statistical physics and circuit layout design. Operations Research36(3), 493–513 (1988) 10

    1988

  5. [5]

    Physical Chemistry Chemical Physics7(23), 3910–3916 (2005)

    Earl, D.J., Deem, M.W.: Parallel tempering: Theory, applications, and new perspectives. Physical Chemistry Chemical Physics7(23), 3910–3916 (2005)

    2005

  6. [6]

    Journal of optimization theory and applications45(1), 41–51 (1985)

    ˇCern` y, V.: Thermodynamical approach to the traveling salesman problem: An efficient simulation algorithm. Journal of optimization theory and applications45(1), 41–51 (1985)

    1985

  7. [7]

    Journal of scheduling7(6), 441–499 (2004)

    Burke, E.K., De Causmaecker, P., Berghe, G.V., Van Landeghem, H.: The state of the art of nurse rostering. Journal of scheduling7(6), 441–499 (2004)

    2004

  8. [8]

    Nature473(7346), 194–198 (2011)

    Johnson, M.W., Amin, M.H., Gildert, S., Lanting, T., Hamze, F., Dickson, N., Harris, R., Berkley, A.J., Johansson, J., Bunyk, P.,et al.: Quantum annealing with manufactured spins. Nature473(7346), 194–198 (2011)

    2011

  9. [9]

    In: 2007 IEEE International Electron Devices Meeting, pp

    Nishiguchi, K., Fujiwara, A.: Single-electron circuit for stochastic data processing using nano- mosfets. In: 2007 IEEE International Electron Devices Meeting, pp. 791–794 (2007). IEEE

    2007

  10. [10]

    Science advances2(6), 1600236 (2016)

    Mahboob, I., Okamoto, H., Yamaguchi, H.: An electromechanical ising hamiltonian. Science advances2(6), 1600236 (2016)

    2016

  11. [11]

    Scientific reports7(1), 44370 (2017)

    Sutton, B., Camsari, K.Y., Behin-Aein, B., Datta, S.: Intrinsic optimization using stochastic nanomagnets. Scientific reports7(1), 44370 (2017)

    2017

  12. [12]

    Nature Electronics6(12), 1009–1019 (2023)

    Whitehead, W., Nelson, Z., Camsari, K.Y., Theogarajan, L.: Cmos-compatible ising and potts annealing using single-photon avalanche diodes. Nature Electronics6(12), 1009–1019 (2023)

    2023

  13. [13]

    Communications Physics6(1), 227 (2023)

    Litvinenko, A., Khymyn, R., Gonz´ alez, V.H., Ovcharov, R., Awad, A.A., Tyberkevych, V., Slavin, A., ˚Akerman, J.: A spinwave ising machine. Communications Physics6(1), 227 (2023)

    2023

  14. [14]

    Applied Physics Letters124(9) (2024)

    Gonz´ alez, V.H., Litvinenko, A., Khymyn, R., ˚Akerman, J.: Global biasing using a hardware- based artificial zeeman term in spinwave ising machines. Applied Physics Letters124(9) (2024)

    2024

  15. [15]

    Current Opinion in Solid State and Materials Science 31, 101173 (2024)

    Gonz´ alez, V.H., Litvinenko, A., Kumar, A., Khymyn, R., ˚Akerman, J.: Spintronic devices as next-generation computation accelerators. Current Opinion in Solid State and Materials Science 31, 101173 (2024)

    2024

  16. [16]

    Nature Communications15(1), 3457 (2024)

    Si, J., Yang, S., Cen, Y., Chen, J., Huang, Y., Yao, Z., Kim, D.-J., Cai, K., Yoo, J., Fong, X.,et al.: Energy-efficient superparamagnetic ising machine and its application to traveling salesman problems. Nature Communications15(1), 3457 (2024)

    2024

  17. [17]

    arXiv preprint arXiv:2406.07197 (2024)

    Ovcharov, R.V., Gonz´ alez, V.H., Litvinenko, A., ˚Akerman, J., Khymyn, R.S.: A numeri- cal model for time-multiplexed ising machines based on delay-line oscillators. arXiv preprint arXiv:2406.07197 (2024)

    work page arXiv 2024

  18. [18]

    npj Quantum Information3(1), 49 (2017)

    Yamamoto, Y., Aihara, K., Leleu, T., Kawarabayashi, K.-i., Kako, S., Fejer, M., Inoue, K., Takesue, H.: Coherent ising machines—optical neural networks operating at the quantum limit. npj Quantum Information3(1), 49 (2017)

    2017

  19. [19]

    Science advances7(40), 0952 (2021)

    Honjo, T., Sonobe, T., Inaba, K., Inagaki, T., Ikuta, T., Yamada, Y., Kazama, T., Enbutsu, K., Umeki, T., Kasahara, R.,et al.: 100,000-spin coherent ising machine. Science advances7(40), 0952 (2021)

    2021

  20. [20]

    Science Advances11(7), 7223 (2025)

    Takesue, H., Inaba, K., Honjo, T., Yamada, Y., Ikuta, T., Yonezu, Y., Inagaki, T., Umeki, T., Kasahara, R.: Finding independent sets in large-scale graphs with a coherent ising machine. Science Advances11(7), 7223 (2025)

    2025

  21. [21]

    Science354(6312), 603–606 (2016)

    Inagaki, T., Haribara, Y., Igarashi, K., Sonobe, T., Tamate, S., Honjo, T., Marandi, A., McMa- hon, P.L., Umeki, T., Enbutsu, K.,et al.: A coherent ising machine for 2000-node optimization problems. Science354(6312), 603–606 (2016)

    2000

  22. [22]

    Communications Physics8(1), 1–11 (2025) 11

    Litvinenko, A., Khymyn, R., Ovcharov, R., ˚Akerman, J.: A 50-spin surface acoustic wave ising machine. Communications Physics8(1), 1–11 (2025) 11

    2025

  23. [23]

    (ed.): RF Bulk Acoustic Wave Filters for Communications

    Hashimoto, K.-y. (ed.): RF Bulk Acoustic Wave Filters for Communications. Artech House, Norwood, MA (2009)

    2009

  24. [24]

    Micromachines11(7), 630 (2020)

    Liu, Y., Cai, Y., Zhang, Y., Tovstopyat, A., Liu, S., Sun, C.: Materials, design, and characteristics of bulk acoustic wave resonator: A review. Micromachines11(7), 630 (2020)

    2020

  25. [25]

    Sensors22(3), 820 (2022)

    Mandal, D., Banerjee, S.: Surface acoustic wave (saw) sensors: Physics, materials, and applications. Sensors22(3), 820 (2022)

    2022

  26. [26]

    Sensor Review44(3), 249–266 (2024)

    Tang, Z., Wu, W., Yang, P., Luo, J., Fu, C., Han, J.-C., Zhou, Y., Wang, L., Wu, Y., Huang, Y.: A review of surface acoustic wave sensors: Mechanisms, stability and future prospects. Sensor Review44(3), 249–266 (2024)

    2024

  27. [27]

    Electrotechnica & Electronica (E+ E)51(2016)

    Terzieva, M.D.: Overview of bulk acoustic wave technology and its applications. Electrotechnica & Electronica (E+ E)51(2016)

    2016

  28. [28]

    In: 2018 IEEE International Electron Devices Meeting (IEDM), pp

    Aigner, R., Fattinger, G., Schaefer, M., Karnati, K., Rothemund, R., Dumont, F.: Baw filters for 5g bands. In: 2018 IEEE International Electron Devices Meeting (IEDM), pp. 14–5 (2018). IEEE

    2018

  29. [29]

    www-user.tu-chemnitz.de/˜helmberg/rudy.tar.gz (1996)

    Rinaldi, G.: rudy graph generator. www-user.tu-chemnitz.de/˜helmberg/rudy.tar.gz (1996)

    1996

  30. [30]

    SIAM Journal on Optimization10(3), 673–696 (2000)

    Helmberg, C., Rendl, F.: A spectral bundle method for semidefinite programming. SIAM Journal on Optimization10(3), 673–696 (2000)

    2000

  31. [31]

    Science advances5(4), 2372 (2019)

    Goto, H., Tatsumura, K., Dixon, A.R.: Combinatorial optimization by simulating adiabatic bifurcations in nonlinear hamiltonian systems. Science advances5(4), 2372 (2019)

    2019

  32. [32]

    https://github.com/bqth29/simulated-bifurcation-algorithm (2025)

    Ageron, R., Bouquet, T., Pugliese, L.: Simulated Bifurcation (SB) algorithm for Python. https://github.com/bqth29/simulated-bifurcation-algorithm (2025)

    2025

  33. [33]

    Science Advances7(6), 7953 (2021)

    Goto, H., Endo, K., Suzuki, M., Sakai, Y., Kanao, T., Hamakawa, Y., Hidaka, R., Yamasaki, M., Tatsumura, K.: High-performance combinatorial optimization based on classical mechanics. Science Advances7(6), 7953 (2021)

    2021

  34. [34]

    Communications Physics5(1), 153 (2022)

    Kanao, T., Goto, H.: Simulated bifurcation assisted by thermal fluctuation. Communications Physics5(1), 153 (2022)

    2022

  35. [35]

    https://www.teledynedefenseelectronics.com/wireless/Wireless Brochures/BAW Devices Product Selection Guide.pdf

    Teledyne: BAW Devices Product Selection Guide. https://www.teledynedefenseelectronics.com/wireless/Wireless Brochures/BAW Devices Product Selection Guide.pdf

  36. [36]

    (eds.) Reducibility among Combinatorial Problems, pp

    Karp, R.M.: In: Miller, R.E., Thatcher, J.W., Bohlinger, J.D. (eds.) Reducibility among Combinatorial Problems, pp. 85–103. Springer, Boston, MA (1972)

    1972

  37. [37]

    Edward Grady, Lueker, G.S.: Probabilistic Analysis of Packing and Partitioning Algorithms

    Coffman, J. Edward Grady, Lueker, G.S.: Probabilistic Analysis of Packing and Partitioning Algorithms. Wiley–Interscience, New York; Chichester (1991)

    1991

  38. [38]

    SIAM Journal on Computing21(1), 59–64 (1992)

    Tsai, L.-H.: Asymptotic analysis of an algorithm for balanced parallel processor scheduling. SIAM Journal on Computing21(1), 59–64 (1992)

    1992

  39. [39]

    IEEE transactions on Information Theory24(5), 525–530 (1978)

    Merkle, R., Hellman, M.: Hiding information and signatures in trapdoor knapsacks. IEEE transactions on Information Theory24(5), 525–530 (1978)

    1978

  40. [40]

    American Scientist90(2), 113–117 (2002)

    Hayes, B.: Computing science: The easiest hard problem. American Scientist90(2), 113–117 (2002)

    2002

  41. [41]

    Journal of Heuristics31(2), 21 (2025)

    Li, Z., Seidel, T., Leib, D., Bortz, M., Heese, R.: Efficient solution of the number partitioning problem on a quantum annealer: a hybrid quantum-classical decomposition approach. Journal of Heuristics31(2), 21 (2025)

    2025

  42. [42]

    Physics Letters A56(5), 421–422 12 (1976)

    Mattis, D.: Solvable spin systems with random interactions. Physics Letters A56(5), 421–422 12 (1976)

    1976

  43. [43]

    Nature communications7(1), 11524 (2016)

    Graß, T., Ravent´ os, D., Juli´ a-D´ ıaz, B., Gogolin, C., Lewenstein, M.: Quantum annealing for the number-partitioning problem using a tunable spin glass of ions. Nature communications7(1), 11524 (2016)

    2016

  44. [44]

    Technical report, Department of Computer Science, The University of Auckland, New Zealand (2014)

    Resnick, T.: Sudoku at the intersection of classical and quantum computing. Technical report, Department of Computer Science, The University of Auckland, New Zealand (2014)

    2014

  45. [45]

    In: Proceedings of the Genetic and Evolutionary Computation Conference Companion, pp

    M¨ ucke, S.: A simple qubo formulation of sudoku. In: Proceedings of the Genetic and Evolutionary Computation Conference Companion, pp. 1958–1962 (2024)

    1958

  46. [46]

    IEEE Transactions on Artificial Intelligence (2025)

    Pignari, R., Fra, V., Macii, E., Urgese, G.: Efficient solution validation of constraint satisfac- tion problems on neuromorphic hardware: the case of sudoku puzzles. IEEE Transactions on Artificial Intelligence (2025)

    2025

  47. [47]

    Physica D: Nonlinear Phenomena, 134809 (2025)

    Shukla, A., Erementchouk, M., Mazumder, P.: Non-binary dynamical ising machines for combinatorial optimization. Physica D: Nonlinear Phenomena, 134809 (2025)

    2025

  48. [48]

    Scientific reports2(1), 1–8 (2012)

    Ercsey-Ravasz, M., Toroczkai, Z.: The chaos within sudoku. Scientific reports2(1), 1–8 (2012)

    2012

  49. [49]

    In: AI&M (2006)

    Lynce, I., Ouaknine, J.: Sudoku as a sat problem. In: AI&M (2006)

    2006

  50. [50]

    IEICE transactions on fundamentals of electronics, communications and computer sciences86(5), 1052–1060 (2003)

    Yato, T., Seta, T.: Complexity and completeness of finding another solution and its application to puzzles. IEICE transactions on fundamentals of electronics, communications and computer sciences86(5), 1052–1060 (2003)

    2003

  51. [51]

    https://minatoyuichiro.medium.com/sudoku-solving-the- worlds-most-difficult-sudoku-problem-using-ising-model-on-javascript-c9b5add0e5c0 13

    Minato, Y.: Solving the world’s most difficult SUDOKU problem using ising model on javascript — minatoyuichiro.medium.com. https://minatoyuichiro.medium.com/sudoku-solving-the- worlds-most-difficult-sudoku-problem-using-ising-model-on-javascript-c9b5add0e5c0 13