Pith. sign in

REVIEW 3 major objections 9 minor 55 references

Spatial Optical Simulator for Classical Statistical Models

T0 review · 3 major / 9 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A single digital micromirror device encodes arbitrary classical spins and reads out the Hamiltonian energy of a spin configuration directly from a camera's central intensity.

desk verdict First SPIM-type simulator for clock, XY, Potts, and Heisenberg models, with clean algebra and honest data, but the optical energy readout itself is never directly calibrated. read the letter →

arxiv 2412.13476 v1 pith:2VYBTE6X submitted 2024-12-18 physics.optics cond-mat.dis-nncond-mat.stat-mechphysics.app-ph

Song-Tao Yu , Ming-Gen He , Sheng Fang , Youjin Deng , Zhen-Sheng Yuan This is my paper · ORCID
classification physics.opticscond-mat.dis-nncond-mat.stat-mechphysics.app-ph MSC 82B2082B2682B80 PACS 42.79.Hp05.50.+q75.10.Hk
keywords spatialopticalsimulatordigitalmicromirrordevicesuperpixelencodingclockmodelXYPottsHeisenbergspinglass
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper claims that a single digital micromirror device can serve as a spatial optical simulator for the clock, XY, Potts, and Heisenberg models. Spins are encoded as complex light fields using 4x4 superpixel patches, and the energy of any spin configuration is read out directly from the intensity at the center of the camera's Fourier plane. Using a feedback loop with Metropolis-Hastings updates, the authors report observing paramagnetic, ferromagnetic, and spin-glass phase behavior on a fully connected network of 100 spins, in agreement with Monte Carlo simulations. If correct, the result extends the reach of spatial optical simulators from the Ising model to a wide class of statistical systems with distinct symmetries.

What carries the argument

The load-bearing element is the DMD superpixel, a 4x4 block of binary mirrors whose on/off pattern produces one of 6561 discrete complex fields $\xi e^{i\phi}$. The central identity is the Fourier-plane center intensity formula $I = \sum_{mn} \xi_m \xi_n e^{i(\phi_m - \phi_n)}$, which converts the pairwise Hamiltonian into a measurable optical intensity and gives the energy for the clock and XY models; sequential DMD patterns extend the same readout to Potts and Heisenberg interactions. This identity is what turns a camera reading into a statistical-mechanics energy.

What would settle it

Rotate a fully ordered ferromagnetic XY ground state by a common angle and check that the measured central intensity is unchanged to within detection noise; any systematic change would show the readout violates the $U(1)$ symmetry the Hamiltonian is supposed to have. The paper itself reports a visible low-temperature discrepancy for the Heisenberg model, so the same test can be applied there to determine whether discretization errors dominate.

Watch

Extended reading notes

Core claim

The central claim is that a 4x4 superpixel of binary DMD mirrors can be programmed to produce a light field with arbitrary complex amplitude $\xi e^{i\phi}$, so any discrete or continuous spin can be encoded by one or more superpixels. The intensity at the center of the back focal plane, $I = \sum_{mn} \xi_m \xi_n e^{i(\phi_m - \phi_n)}$, equals the pairwise interaction sum of the target Hamiltonian up to an additive constant. For clock and XY models the phase $\phi$ directly represents the spin angle; for Potts spins the interaction is mapped onto vectors so that the Kronecker delta becomes a vector product; for Heisenberg spins the three components are encoded in three consecutive DMD images. Reading $I$ for each image sequence thus yields the energy, and the scheme preserves the $Z_q$, $U(1)$, $S_q$, and $O(3)$ symmetries of the respective models.

Load-bearing premise

The entire simulator rests on the assumption that each superpixel produces its target complex field faithfully enough that the camera's central intensity is a clean coherent sum, with negligible crosstalk, phase drift, and detection nonlinearity.

Editorial extensions

If this is right

  • The same DMD platform can handle any model whose Hamiltonian is a sum of pairwise products of encoded spin variables, including many-body interaction terms.
  • Because the energy readout is optically parallel, the approach inherits the linear-in-$N$ complexity of spatial photonic Ising machines, avoiding the $O(N^2)$ pairwise cost.
  • The observed replica order-parameter peaks in the 3-state clock model indicate that the device can probe multivalley energy landscapes and spin-glass phases, not just ferromagnetic order.
  • Higher-state clock and Potts models can be simulated within the limits of the superpixel's discrete field set, and the reported agreement with Monte Carlo holds for the simulated temperatures and system sizes.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural extension the authors do not spell out is to read off correlations from off-center diffraction intensities, which would extract more than the total energy from the same optical field.
  • Replacing the 4x4 superpixel with larger patches should systematically reduce the discretization errors the paper concedes at low temperature in the Heisenberg model, a testable prediction of the encoding-error explanation.
  • The same encoding strategy could represent complex-valued couplings directly by adding a reference phase to the superpixel pattern, avoiding the need for separate amplitude-only coupling masks.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 9 minor

Summary. The manuscript reports a DMD-based 'spatial optical simulator' for four classical statistical models - the clock, XY, Potts, and Heisenberg models - on fully connected graphs. Spins are encoded by 4x4 mirror superpixels that supply a discrete library of complex fields, and the intensity at the center of the Fourier plane is argued to equal sum_{mn} xi_m xi_n exp[i(phi_m - phi_n)] (Eq. 2), so that the Hamiltonian is read out directly from the camera. The Potts delta interaction is mapped onto a vector dot product via the identity delta_{sm,sn} = (S_m.S_n - q + 4)/4, and the Heisenberg interaction is assembled from three component images with 0/pi phases and amplitudes |S^k|. A computer-DMD-CCD feedback loop performs Metropolis sampling with simulated annealing and global symmetry moves, evaluating the energy optically at each proposal. For N = 100 spins, the authors report energies, magnetizations, and order-parameter distributions for ferromagnetic couplings (q = 3, 4, 8 clock; XY; q = 2, 3, 6 Potts; Heisenberg) and for Mattis-type random couplings on the 3-state clock, and they claim observation of the corresponding ferromagnetic and spin-glass transitions in agreement with digital Monte Carlo simulations.

Significance. If the residual concerns are resolved, this is a clean proof of principle that a single DMD with the superpixel technique can encode vector-valued classical spins (Z_q, U(1), S_q, O(3)) and evaluate their energies through a single-pixel intensity measurement. The algebraic core is sound and parameter-free: Eq. (2) is the standard coherent-superposition relation, the Potts identity checks out, and the temperature normalization in the acceptance rule is internally consistent with the mean-field T_c values in Table I (verified for the q = 2 Potts, XY, and Heisenberg cases). The comparisons use independent digital MC with no fitted parameters, so there is no circularity in the agreement; however, neither the MC comparison nor the reported error bars (which reflect sample correlation only) substitutes for a direct test of the optical readout. The increment over the established SPIM literature (Refs. [18,21,22]) is the multi-model encoding, which is elegant but modest, so the value of the paper rests largely on the experimental fidelity of Eq. (2), the very point that is not yet validated.

major comments (3)
  1. [Eq. (2) and the Metropolis loop (Optical simulation of statistical models)] The most load-bearing element of the manuscript is Eq. (2), which equates the CCD central intensity with sum_{mn} xi_m xi_n exp(i(phi_m - phi_n)). This equation feeds directly into the acceptance rule p = min(1, exp(-Delta H/(2N T))), so any systematic readout error biases the sampled ensemble. The concern raised in review that Eq. (2) is never directly validated is, on my reading of the paper, valid: there is no calibration of the 6561-field superpixel library on the DLP3000 at 785 nm, no interferometric or modulation-depth measurement of the complex amplitudes actually delivered by the superpixels, no direct comparison of the measured central intensity with the digitally computed value of sum_{mn} xi_m xi_n cos(phi_m - phi_n) for a set of known test configurations, and no characterization of CCD linearity, dark signal, or stray-light background. The agreement with MC in Figs. 2-4 is suggestive but does not isolate Eq. (2), because both pipelines share the same Hamiltonian, annealing protocol, and update scheme, and the reported error bars account only for sample correlation. I request an explicit validation (for example, measured versus computed intensity over many random configurations, with a residual analysis) and an error budget demonstrating that the uncertainty in Delta H is small compared with the scale of the acceptance threshold.
  2. [Fig. 2(c) and the following paragraph] The abstract and the conclusion claim quantitative agreement with Monte Carlo simulations, but Fig. 2(c) shows an unquantified low-temperature deviation for the Heisenberg model, which the text attributes to 'discretization encoding errors, detection noise, or aberrations.' This is load-bearing because the Heisenberg sector is the strictest test of the vector-embedding scheme and because none of the three proposed causes is quantified. Please (i) quantify the superpixel library errors for the amplitude encoding of |S^k| and the sign, (ii) estimate the resulting bias in the effective Hamiltonian actually sampled, including the interplay with the enforced normalization (S^x)^2+(S^y)^2+(S^z)^2 = 1, and (iii) show that the bias does not shift the inferred T_c beyond the accuracy claimed in Table I, or restrict the quantitative-agreement claim to the clock, XY, and Potts sectors.
  3. [Fig. 4 and following paragraph (Mattis-type spin glass)] The claim that the Mattis-type 3-state clock exhibits a spin-glass phase with a replica-symmetry-breaking nature is not established by the evidence presented. For q = 2 the Mattis coupling is gauge-equivalent to a ferromagnet and has no spin-glass phase; for q = 3 the model is genuinely frustrated because a bond with xi_m xi_n = -1 cannot be satisfied within the discrete clock phases, but the manuscript gives neither this argument nor any diagnostic that distinguishes glassy ordering from the discrete-symmetry ferromagnetic order of the same model without disorder. A multi-peaked P(Q) is not by itself discriminating, since a ferromagnetic 3-state clock also yields peaks at the values cos(2 pi k/3). I ask for additional diagnostics (for example, the Edwards-Anderson overlap, a comparison with the p = 0 ferromagnetic case, or a statement of the peak positions and their interpretation) or a softening of the 'spin-glass' and 'replica-symmetry-breaking' language.
minor comments (9)
  1. [Fig. 4 paragraph] The definition of the replica order parameter contains a typographical index error: Q_{alpha beta} = (1/N) sum cos(phi^alpha_m - phi^beta_n) should have phi^beta_m as the second argument of the cosine.
  2. [Optical encoding of statistical models] The sentence 'An additional DMD image can be employed to encode a set of nonuniform coupling strength as J_mn = xi_m xi_n' is unclear, because in the preceding description the coupling is set by the superpixel amplitudes in the same image; please clarify how an extra image introduces pairwise products without modifying the spin encoding, or remove the sentence since the reported experiments use only uniform and Mattis couplings.
  3. [Experimental setup, Fig. 1(a)] Please specify which diffraction order of the DMD is collected at the 'center position' of the back focal plane and how the other orders are suppressed; Eq. (2) is valid only for the field in a single selected order, and for the DC component of a binary pattern the superpixel phase would not be controllable.
  4. [Optical simulation of statistical models] The sampling protocol needed for reproducibility is not fully specified: please report the annealing schedule, the number of Metropolis sweeps and camera exposures per temperature, the number of samples per temperature, and the specific method used to estimate error bars (beyond 'taking into account the correlation between samples').
  5. [Table I] Please state or reference the mean-field calculation for the theoretical T_c values and note explicitly that they correspond to the 2NT normalization in the acceptance rule, so that a reader can connect Table I to the curves in Fig. 2.
  6. [Fig. 3 caption] The sentence 'The deviation of the histogram originates from the camera detection noise and encoding errors' is an assertion without supporting data; if no quantitative estimate is available, it should be rephrased as a hypothesis or supported by the calibration requested above.
  7. [Optical simulation of statistical models, Fig. 2] For N = 100 the quantities in Fig. 2 are rounded crossover curves rather than sharp transitions; please state how the experimental T_c was assigned from the data (for example, by the inflection point of the energy density) and how the expected finite-size rounding at N = 100 affects the comparison with Table I.
  8. [Abstract and Optical encoding of statistical models] The abstract's 'precisely encoded' and the statement that 'more than a hundred grayscale levels' allow q to 'reach the hundreds' are stronger than the 6561-field discretization supports; a hundred phase levels correspond to q up to about one hundred for the clock model.
  9. [Paragraph after Fig. 2] In the sentence discussing the Heisenberg discrepancies, the parenthetical reference to Fig. 2(a) appears to be a citation error; the low-temperature Heisenberg data are in Fig. 2(c).

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the optical Hamiltonian mapping is constructive and the phase-transition results are benchmarked against independent Monte Carlo simulations.

full rationale

Walking the derivation chain, the optical energy readout is not circular: the target Hamiltonians in Eq. (1) are independently defined, and the paper then constructs an encoding (single superpixel for clock/XY, q superpixels for Potts, three superpixels for Heisenberg) and shows by direct algebra that the CCD central intensity in Eq. (2) equals the corresponding Hamiltonian up to a constant. That is a constructive mapping, not a fit or a renamed prediction: H_clock = -I, the Potts identity delta_{sigma_m,sigma_n} = (S_m dot S_n - q + 4)/4, and the Heisenberg relation I = sum S_m dot S_n are algebraic identities under the stated encoding. The experimental claims (phase transitions, magnetization, and the spin-glass replica order parameter) are compared against Monte Carlo simulations on a digital computer, which are independent of the optical apparatus, and no parameter is adjusted to force agreement. The only self-referential element is Ref. [45], co-authored by two of the present authors, used for the q >= 5 clock-model universality statement; this is corroborated by Refs. [43] and [44], is not load-bearing for the optical construction, and does not forbid alternatives. The absence of direct calibration of the 6561 superpixel fields or detector linearity is a correctness-verification gap, not circularity: the readout equation is assumed rather than experimentally proven, but that assumption is not equivalent to the target result by construction. Hence no circular step is present.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The central claim rests on standard Fourier optics, the known 4x4 superpixel encoding technique, elementary spin-vector embeddings, and textbook Monte Carlo sampling. No new physical entities are introduced and no constants are fitted to the benchmark data. The main unexamined premise is the quantitative fidelity of the DMD superpixel encoding, which the authors partially acknowledge through low-temperature discrepancies.

assumptions (6)
  • domain assumption A 4x4 superpixel can encode any complex field value xi e^{i phi} with xi in [0,1] and phi in [0,2 pi].
    Invoked in 'Optical encoding of statistical models'; relies on Goorden et al. (ref 42). The paper acknowledges only 6561 discrete fields, so this is approximate.
  • domain assumption The CCD intensity at the back focal plane equals the modulus-square of the coherent sum of superpixel fields (Eq. 2).
    Standard Fourier optics assumption; central to energy readout.
  • standard math delta_{sigma_m, sigma_n} = (S_m . S_n - q + 4) / 4 for the +/-1 Potts embedding.
    Algebraic identity stated in text; verified for equal and different spins.
  • domain assumption For q >= 5 the q-state clock model transitions are in the XY universality class.
    Cited from refs 43-45; used to interpret q=8 and XY results.
  • standard math Metropolis-Hastings acceptance with probability min{1, e^{-Delta H / (2 N T)}} samples the Boltzmann distribution.
    Standard algorithm (ref 53); the 2N factor accounts for the fully connected double-counted Hamiltonian normalization.
  • domain assumption The p=0.5 Mattis-type random interaction model exhibits a spin-glass phase.
    Assumed when interpreting multiple peaks of Q as spin-glass order; supported by Mattis/Nishimori literature and the paper's own MC comparison.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Spatial Optical Simulator for Classical Statistical Models." pith.science (2026). https://pith.science/paper/2VYBTE6X

@misc{pith2026241213476,
  author       = {Pith},
  title        = {Pith review of: Spatial Optical Simulator for Classical Statistical Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2VYBTE6X}},
  note         = {Machine review of arXiv:2412.13476}
}
abstract

Optical simulators for the Ising model have demonstrated great promise for solving challenging problems in physics and beyond. Here, we develop a spatial optical simulator for a variety of classical statistical systems, including the clock, $XY$, Potts, and Heisenberg models, utilizing a digital micromirror device composed of a large number of tiny mirrors. Spins, with desired amplitudes or phases of the statistical models, are precisely encoded by a patch of mirrors with a superpixel approach. Then, by modulating the light field in a sequence of designed patterns, the spin-spin interaction is realized in such a way that the Hamiltonian symmetries are preserved. We successfully simulate statistical systems on a fully connected network, with ferromagnetic or Mattis-type random interactions, and observe the corresponding phase transitions between the paramagnetic, and the ferromagnetic or spin-glass phases. Our results largely extend the research scope of spatial optical simulators and their versatile applications.

Figures

Figures reproduced from arXiv: 2412.13476 by the authors.

Figure 1
Figure 1. FIG. 1. Scheme of the DMD-based optical simulators for sta [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. The energy density [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. The spontaneous symmetry breaking in the 8-state [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: FIG. 4. The spin glass in the 3-state clock model. The [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

55 extracted references · 44 canonical work pages

  1. [1]

    H. J. Caulfield and S. Dolev, Why future supercomputing requires optics, Nature Photonics 4, 261 (2010)

    work page 2010

  2. [2]

    H. Zhou, J. Dong, J. Cheng, W. Dong, C. Huang, Y. Shen, Q. Zhang, M. Gu, C. Qian, H. Chen, et al., Photonic matrix multiplication lights up photonic accel- erator and beyond, Light: Science & Applications 11, 30 (2022)

    work page 2022

  3. [3]

    P. L. McMahon, The physics of optical computing, Na- ture Reviews Physics 5, 717 (2023)

    2023

  4. [4]

    Ghofraniha, I

    N. Ghofraniha, I. Viola, F. Di Maria, G. Barbarella, G. Gigli, L. Leuzzi, and C. Conti, Experimental evidence of replica symmetry breaking in random lasers, Nature Communications 6, 6058 (2015)

    work page 2015

  5. [5]

    Vandoorne, P

    K. Vandoorne, P. Mechet, T. Van Vaerenbergh, M. Fiers, G. Morthier, D. Verstraeten, B. Schrauwen, J. Dambre, and P. Bienstman, Experimental demonstration of reser- voir computing on a silicon photonics chip, Nature Com- munications 5, 3541 (2014)

    2014

  6. [6]

    X. Xu, M. Tan, B. Corcoran, J. Wu, A. Boes, T. G. Nguyen, S. T. Chu, B. E. Little, D. G. Hicks, R. Moran- dotti, et al., 11 TOPS photonic convolutional accelerator for optical neural networks, Nature 589, 44 (2021)

    work page 2021

  7. [7]

    X. Lin, Y. Rivenson, N. T. Yardimci, M. Veli, Y. Luo, M. Jarrahi, and A. Ozcan, All-optical machine learning using diffractive deep neural networks, Science 361, 1004 (2018)

    2018

  8. [8]

    Feldmann, N

    J. Feldmann, N. Youngblood, C. D. Wright, H. Bhaskaran, and W. H. Pernice, All-optical spiking neurosynaptic networks with self-learning capabilities, Nature 569, 208 (2019)

    2019

Show all 55 references
  1. [9]

    Wetzstein, A

    G. Wetzstein, A. Ozcan, S. Gigan, S. Fan, D. Englund, M. Soljaˇ ci´ c, C. Denz, D. A. Miller, and D. Psaltis, Infer- ence in artificial intelligence with deep optics and pho- tonics, Nature 588, 39 (2020)

    2020

  2. [10]

    Zhuang, C

    L. Zhuang, C. G. Roeloffzen, M. Hoekman, K.-J. Boller, and A. J. Lowery, Programmable photonic signal proces- sor chip for radiofrequency applications, Optica 2, 854 (2015)

    2015

  3. [11]

    P´ erez, I

    D. P´ erez, I. Gasulla, L. Crudgington, D. J. Thomson, A. Z. Khokhar, K. Li, W. Cao, G. Z. Mashanovich, and J. Capmany, Multipurpose silicon photonics signal pro- cessor core, Nature Communications 8, 636 (2017)

    2017

  4. [12]

    Bogaerts, D

    W. Bogaerts, D. P´ erez, J. Capmany, D. A. Miller, J. Poon, D. Englund, F. Morichetti, and A. Melloni, Pro- grammable photonic circuits, Nature 586, 207 (2020)

    2020

  5. [13]

    S. Pai, T. Park, M. Ball, B. Penkovsky, M. Dubrovsky, 6 N. Abebe, M. Milanizadeh, F. Morichetti, A. Melloni, S. Fan, et al., Experimental evaluation of digitally veri- fiable photonic computing for blockchain and cryptocur- rency, Optica 10, 552 (2023)

    2023

  6. [14]

    Marandi, Z

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

    2014

  7. [15]

    Inagaki, Y

    T. Inagaki, Y. Haribara, K. Igarashi, T. Sonobe, S. Ta- mate, T. Honjo, A. Marandi, P. L. McMahon, T. Umeki, and K. Enbutsu, A coherent Ising machine for 2000-node optimization problems, Science 354, 603 (2016)

    2016

  8. [16]

    Okawachi, M

    Y. Okawachi, M. Yu, J. K. Jang, X. Ji, Y. Zhao, B. Y. Kim, M. Lipson, and A. L. Gaeta, Demonstration of chip- based coupled degenerate optical parametric oscillators for realizing a nanophotonic spin-glass, Nature Commu- nications 11, 4119 (2020)

    2020

  9. [17]

    Roques-Carmes, Y

    C. Roques-Carmes, Y. Shen, C. Zanoci, M. Prabhu, F. Atieh, L. Jing, T. Dubˇ cek, C. Mao, M. R. Johnson, V. ˇCeperi´ c,et al., Heuristic recurrent algorithms for pho- tonic Ising machines, Nature Communications 11, 249 (2020)

    2020

  10. [18]

    Pierangeli, G

    D. Pierangeli, G. Marcucci, and C. Conti, Large-scale photonic Ising machine by spatial light modulation, Physical Review Letters 122, 213902 (2019)

    2019

  11. [19]

    Yamashita, K.-I

    H. Yamashita, K.-I. Okubo, S. Shimomura, Y. Ogura, J. Tanida, and H. Suzuki, Low-rank combinatorial opti- mization and statistical learning by spatial photonic Ising machine, Physical Review Letters 131, 063801 (2023)

    2023

  12. [20]

    Pierangeli, G

    D. Pierangeli, G. Marcucci, and C. Conti, Adiabatic evo- lution on a spatial-photonic Ising machine, Optica 7, 1535 (2020)

    2020

  13. [21]

    Leonetti, E

    M. Leonetti, E. H¨ ormann, L. Leuzzi, G. Parisi, and G. Ruocco, Optical computation of a spin glass dynam- ics with tunable complexity, Proceedings of the National Academy of Sciences 118, e2015207118 (2021)

    2021

  14. [22]

    Pierangeli, M

    D. Pierangeli, M. Rafayelyan, C. Conti, and S. Gigan, Scalable spin-glass optical simulator, Physical Review Applied 15, 034087 (2021)

    2021

  15. [23]

    Kumar, Z

    S. Kumar, Z. Li, T. Bu, C. Qu, and Y. Huang, Observa- tion of distinct phase transitions in a nonlinear optical Ising machine, Communications Physics 6, 31 (2023)

    2023

  16. [24]

    Y. Fang, J. Huang, and Z. Ruan, Experimental obser- vation of phase transitions in spatial photonic Ising ma- chine, Physical Review Letters 127, 043902 (2021)

    2021

  17. [25]

    Mohseni, P

    N. Mohseni, P. L. McMahon, and T. Byrnes, Ising ma- chines as hardware solvers of combinatorial optimization problems, Nature Reviews Physics 4, 363 (2022)

    2022

  18. [26]

    Prabhakar, P

    A. Prabhakar, P. Shah, U. Gautham, V. Natarajan, V. Ramesh, N. Chandrachoodan, and S. Tayur, Opti- mization with photonic wave-based annealers, Philosoph- ical Transactions of the Royal Society A 381, 20210409 (2023)

    2023

  19. [27]

    Zhang and Y

    S. Zhang and Y. Huang, Complex quadratic optimization and semidefinite programming, SIAM Journal on Opti- mization 16, 871 (2006)

    2006

  20. [28]

    A. M.-C. So, J. Zhang, and Y. Ye, On approximating complex quadratic optimization problems via semidefi- nite programming relaxations, Mathematical Program- ming 110, 93 (2007)

    2007

  21. [29]

    Frieze and M

    A. Frieze and M. Jerrum, Improved approximation algo- rithms for MAX k-CUT and MAX BISECTION, Algo- rithmica 18, 67 (1997)

    1997

  22. [30]

    R. M. R. Lewis, A Guide to Graph Colouring: Algorithms and Applications (Spring, Cham, 2015)

    2015

  23. [31]

    Hamerly, K

    R. Hamerly, K. Inaba, T. Inagaki, H. Takesue, Y. Ya- mamoto, and H. Mabuchi, Topological defect formation in 1D and 2D spin chains realized by network of optical parametric oscillators, International Journal of Modern Physics B 30, 1630014 (2016)

    2016

  24. [32]

    Takeda, S

    Y. Takeda, S. Tamate, Y. Yamamoto, H. Takesue, T. Ina- gaki, and S. Utsunomiya, Boltzmann sampling for an XY model using a non-degenerate optical parametric oscilla- tor network, Quantum Science and Technology3, 014004 (2017)

    2017

  25. [33]

    Honari-Latifpour and M.-A

    M. Honari-Latifpour and M.-A. Miri, Optical Potts ma- chine through networks of three-photon down-conversion oscillators, Nanophotonics 9, 4199 (2020)

    2020

  26. [34]

    Inaba, T

    K. Inaba, T. Inagaki, K. Igarashi, S. Utsunomiya, T. Honjo, T. Ikuta, K. Enbutsu, T. Umeki, R. Kasahara, K. Inoue, et al., Potts model solver based on hybrid phys- ical and digital architecture, Communications Physics 5, 137 (2022)

    2022

  27. [35]

    N. G. Berloff, M. Silva, K. Kalinin, A. Askitopoulos, J. D. T¨ opfer, P. Cilibrizzi, W. Langbein, and P. G. Lagoudakis, Realizing the classical XY Hamiltonian in polariton sim- ulators, Nature Materials 16, 1120 (2017)

    2017

  28. [36]

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

    2020

  29. [37]

    K. P. Kalinin and N. G. Berloff, Simulating Ising and n-state planar Potts models and external fields with nonequilibrium condensates, Physical Review Letters 121, 235302 (2018)

    2018

  30. [38]

    S. L. Harrison, H. Sigurdsson, S. Alyatkin, J. D. T¨ opfer, and P. G. Lagoudakis, Solving the max-3-cut prob- lemwith coherent networks, Physical Review Applied17, 024063 (2022)

    2022

  31. [39]

    Nixon, E

    M. Nixon, E. Ronen, A. A. Friesem, and N. Davidson, Observing geometric frustration with thousands of cou- pled lasers, Physical Review Letters 110, 184102 (2013)

    2013

  32. [40]

    Gershenzon, G

    I. Gershenzon, G. Arwas, S. Gadasi, C. Tradonsky, A. Friesem, O. Raz, and N. Davidson, Exact mapping between a laser network loss rate and the classical XY Hamiltonian by laser loss control, Nanophotonics 9, 4117 (2020)

    2020

  33. [41]

    Turtaev, I

    S. Turtaev, I. T. Leite, K. J. Mitchell, M. J. Padgett, D. B. Phillips, and T. ˇCiˇ zm´ ar, Comparison of nematic liquid-crystal and DMD based spatial light modulation in complex photonics, Optics Express 25, 29874 (2017)

    2017

  34. [42]

    S. A. Goorden, J. Bertolotti, and A. P. Mosk, Superpixel- based spatial amplitude and phase modulation using a digital micromirror device, Optics Express 22, 17999 (2014)

    2014

  35. [43]

    Borisenko, G

    O. Borisenko, G. Cortese, R. Fiore, M. Gravina, and A. Papa, Numerical study of the phase transitions in the two-dimensional Z(5) vector model, Physical Review E 83, 041120 (2011)

    2011

  36. [44]

    Kumano, K

    Y. Kumano, K. Hukushima, Y. Tomita, and M. Os- hikawa, Response to a twist in systems with Zp symme- try: The two-dimensional p-state clock model, Physical Review B 88, 104427 (2013)

    2013

  37. [45]

    H. Chen, P. Hou, S. Fang, and Y. Deng, Monte Carlo study of duality and the Berezinskii-Kosterlitz-Thouless phase transitions of the two-dimensional q-state clock model in flow representations, Physical Review E 106, 024106 (2022)

    2022

  38. [46]

    Wu, The Potts model, Reviews of Modern Physics 7 54, 235 (1982)

    F.-Y. Wu, The Potts model, Reviews of Modern Physics 7 54, 235 (1982)

    1982

  39. [47]

    Mattis, Solvable spin systems with random interac- tions, Physics Letters A 56, 421 (1976)

    D. Mattis, Solvable spin systems with random interac- tions, Physics Letters A 56, 421 (1976)

    1976

  40. [48]

    Nishimori, Statistical Physics of Spin Glasses and- Information Processing: An Introduction (Clarendon Press, New York, 2001)

    H. Nishimori, Statistical Physics of Spin Glasses and- Information Processing: An Introduction (Clarendon Press, New York, 2001)

    2001

  41. [49]

    Huang, Y

    J. Huang, Y. Fang, and Z. Ruan, Antiferromagnetic spatial photonic Ising machine through optoelectronic correlation computing, Communications Physics 4, 242 (2021)

    2021

  42. [50]

    L. Luo, Z. Mi, J. Huang, and Z. Ruan, Wavelength- division multiplexing optical Ising simulator enabling fully programmable spin couplings and external magnetic fields, Science Advances 9, eadg6238 (2023)

    2023

  43. [51]

    Sakabe, S

    T. Sakabe, S. Shimomura, Y. Ogura, K.-i. Okubo, H. Ya- mashita, H. Suzuki, and J. Tanida, Spatial-photonic Ising machine by space-division multiplexing with physically tunable coefficients of a multi-component model, Optics Express 31, 44127 (2023)

    2023

  44. [52]

    W. Sun, W. Zhang, Y. Liu, Q. Liu, and Z. He, Quadra- ture photonic spatial Ising machine, Optics Letters 47, 1498 (2022)

    2022

  45. [53]

    W. K. Hastings, Monte Carlo sampling methods using Markov chains and their applications, Biometrika 57, 97 (1970)

    1970

  46. [54]

    Chibante, Simulated Annealing: Theory with Applica- tions (Sciyo, Rijeka, 2010)

    R. Chibante, Simulated Annealing: Theory with Applica- tions (Sciyo, Rijeka, 2010)

    2010

  47. [55]

    Gelman, J

    A. Gelman, J. B. Carlin, H. S. Stern, and D. B. Rubin, Bayesian Data Analysis(Chapman and Hall/CRC, Boca Raton, 2004)

    2004

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.