Pith. sign in

REVIEW 3 major objections 5 minor 63 references

Higher-Order Neuromorphic Ising Machines -- Autoencoders and Fowler-Nordheim Annealers are all you need for Scalability

T0 review · 3 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read Operating on Ising clauses rather than on spins lets a neuromorphic architecture anneal arbitrary-order Hamiltonians with hardware cost that does not grow with interaction order, and the Fowler–Nordheim schedule preserves the asymptotic…

desk verdict The delta-E rewrite and autoencoder framing are sound and worth building on, but the convergence guarantee doesn't cover the solvers that produced the benchmarks—so the headline claim needs either a new proof or a haircut. read the letter →

arxiv 2506.19964 v1 pith:G6AEFKBZ submitted 2025-06-24 cs.NE

Faiek Ahsan , Saptarshi Maiti , Zihao Chen , Jakob Kaiser , Ankita Nandi , Madhuvanthi Srivatsav , Johannes Schemmel , Andreas G. Andreou
show 3 more authors
Jason Eshraghian Chetan Singh Thakur Shantanu Chakrabartty
This is my paper · ORCID
classification cs.NE
keywords NeuromorphicComputingIsingMachinesHigher-orderFowler-NordheimAnnealingAutoencoderMAX-SATMAX-CUTGraphColoring
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 tries to establish that a neuromorphic Ising machine can solve higher-order combinatorial optimization problems—MAX-SAT with three-, five-, and seven-body clauses—without first quadratizing them into pairwise interactions. The architecture makes the dynamic variables the clauses themselves, $\prod_i s_i^{H_{k,i}}$, while sampling the energy landscape in the lower-dimensional spin space through an autoencoder, so hardware resources scale with the clause–variable incidence structure rather than with interaction order or auxiliary-variable overhead. The same Fowler–Nordheim annealer that guarantees asymptotic convergence for pairwise Ising machines is shown to transfer to this setting, because flipping a spin lowers the energy exactly when the weighted clause sum feeding that spin is negative. If the claim holds, quadratization ceases to be the bottleneck for higher-order Ising hardware, and even pairwise problems like MAX-CUT can be solved at time-to-solution competitive with specialized solvers.

What carries the argument

The load-bearing object is the clause–spin autoencoder defined by the binary incidence matrix $H\in\{0,1\}^{M\times N}$: the encoder maps the $M$ clause outputs $T_k$ to $N$ latent spin neurons through the weighted sum $\sum_k H_{k,i} J_k T_k$, and the decoder reconstructs each clause from the parity of latent-neuron spikes. The identity carrying the argument is $\Delta E_i<0 \iff \sum_k H_{k,i} J_k T_{k,n-1}<0$, which turns the energy comparison of simulated annealing into a thresholded weighted sum. The Fowler–Nordheim annealer supplies the time-varying noisy thresholds $\mu_{i,n}=\tau_n N^E_n + A N^B_{i,n}$ with $\tau_n\sim 1/\log n$, reproducing the optimal annealing schedule and, in the idealized asynchronous model, preserving asymptotic convergence. Decoder toggle neurons implement clause flips, and graph coloring of latent neurons allows conditionally independent spins to update in parallel.

What would settle it

Take a planted 3R-3X instance with a known ground state, run the graph-colored implementation of SI Algorithm 2 with the Bernoulli threshold term removed, as in the FPGA results, and check whether it always reaches the planted ground state under the Fowler–Nordheim schedule. If any instance gets stuck in a local minimum that the asynchronous Bernoulli-threshold dynamics escapes, the claim that the implemented variants inherit the asymptotic convergence guarantee is false.

Watch

Extended reading notes

Core claim

The central discovery is that the general higher-order Ising Hamiltonian $E(s)=-\sum_{k=1}^{M} J_k \prod_{i=1}^{N} s_i^{H_{k,i}}$ admits an autoencoder implementation whose outer layer stores the clause outputs $T_k=\prod_i s_i^{H_{k,i}}$ and whose latent layer stores the spins $s_i$. Flipping spin $i$ lowers the energy if and only if $\sum_{k=1}^{M} H_{k,i} J_k T_{k,n-1} < 0$; the encoder computes that weighted clause sum at each latent neuron, the noisy threshold implements the simulated-annealing acceptance step, and the decoder toggles a clause whenever the parity of contributing spin flips is odd. Because clauses are manipulated directly, a $p$-body interaction costs one row of the interconnection matrix instead of $O(n^p)$ pairwise couplings, and the auxiliary variables of quadratization disappear. With the Fowler–Nordheim temperature schedule $\tau(t)=\tau_0/\log(1+t/C)$, the paper argues the system asymptotically converges to the Ising ground state. On MAX-CUT and MAX-3SAT benchmarks, the higher-order machine yields better solutions in fewer iterations than the same annealer applied to a quadratized second-order model; graph coloring and an FPGA implementation further improve time-to-solution.

Load-bearing premise

The convergence guarantee is derived for asynchronous single-spin-flip updates with Bernoulli noise, but the benchmark results come from rejection-free parallel selection and graph-colored simultaneous updates; if those variants do not preserve the same sampling statistics, the reported performance is a heuristic rather than a consequence of the guarantee.

Editorial extensions

If this is right

  • Any-order MAX-SAT clauses can be embedded directly, so the search space is not inflated by auxiliary variables; on the paper's benchmarks this yields near-perfect satisfiability at sizes where the quadratized second-order solver's success rate collapses.
  • For sparse problem encodings the hardware footprint becomes $O(nq)$ for the encoder and $O(mp)$ for the decoder instead of $O(n^p)$ couplings or quadratization's exponential auxiliary-variable overhead.
  • Graph coloring massively parallelizes updates of conditionally independent latent neurons, accelerating high-temperature convergence; sufficiently dense graphs must fall back to the uncolored asynchronous variant to preserve ergodicity.
  • An FPGA implementation with 16-bit noise thresholds reaches 99% clause satisfaction on size-250 MAX-3SAT instances in median times under a microsecond, a 2500–3000x speedup over the CPU emulation at the same precision.
  • On the second-order MAX-CUT benchmark the same machine matches or beats the reported state-of-the-art cuts on G4, G11, and G15 with competitive time-to-solution.

Reading between the lines

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

  • If the resource claim holds at scale, the real scaling ceiling shifts to clause count and clause density; a clean test would hold the clause count fixed while raising the interaction order and measure whether encoder and decoder resources stay flat.
  • The convergence guarantee is proven for the asynchronous single-spin-flip model, while the reported runs use rejection-free and graph-colored parallel updates; a reader should treat the guarantee as covering the idealized dynamics until the variants are proven equivalent.
  • Because a clause output is a parity product of spins, the same encoder–decoder factorization could serve parity-check decoding problems beyond SAT, such as LDPC decoding, where higher-order parity constraints are native and quadratization is also unnatural.
  • A physical Fowler–Nordheim tunneling device is the natural next noise source; the paper notes digital emulation was chosen for 16-bit precision, so a device-based implementation would test whether the annealer's hardware promise survives outside simulation.
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 / 5 minor

Summary. The paper proposes a neuromorphic Ising machine for higher-order Hamiltonians of the form Eq. (1), organized as an autoencoder whose outer layer consists of clause neurons and whose latent layer consists of spin neurons. The central algebraic insight is that the energy change on flipping spin i can be expressed as 2∑_k H_{k,i} J_k T_{k,n-1} (Eqs. (7)-(10)), so that the Metropolis acceptance test of simulated annealing can be implemented by comparing this weighted clause sum against a noisy threshold. The noisy threshold is generated by an exponentially distributed random variable annealed with a Fowler-Nordheim schedule τ(t)=τ0/log(1+t/C) (Eq. (19)), plus (in the idealized model) a Bernoulli term to enforce asynchrony. The paper claims asymptotic convergence to the Ising ground state by exact functional isomorphism to simulated annealing, and claims resource complexity independent of interaction order. It reports benchmarks on MAX-CUT (Gset), MAX-3SAT (SATLIB), and MAX-5/7SAT instances, with CPU and FPGA implementations, comparing against the second-order NeuroSA machine and the dSBM simulator.

Significance. If the convergence and resource claims were established for the actual implementations, this would be a significant contribution: the algebraic reformulation in Eq. (2)/(10) is exact and clean, the mapping of the Metropolis accept/reject step to thresholded latent neurons (Eqs. (13)-(16)) is an elegant and genuinely useful rewrite for the asynchronous single-spin-flip model, and the paper ships extensive benchmark comparisons, including 1000-trial MAX-CUT runs, FPGA timing measurements, and comparisons to a commercial-style simulator (dSBM). The claim of avoiding quadratization overhead for higher-order interactions is of broad interest to the Ising-machine and neuromorphic-computing communities. However, the headline convergence guarantee is proven only for the idealized asynchronous Bernoulli model, while the solvers that produced all reported results use different update rules; this gap is central to the paper's reliability claims and must be addressed before the significance can be accepted.

major comments (3)
  1. The asymptotic convergence guarantee is established only for the asynchronous, Bernoulli-threshold update of Eqs. (14)-(16), which is an exact rewrite of Metropolis simulated annealing. The implementations that produced every reported figure — SI Algorithm 1 (uncolored) and SI Algorithm 2 (graph-colored) — drop the Bernoulli term N^B and instead perform parallel thresholding followed by either uniform random selection among all active neurons (Algorithm 1) or simultaneous flips of all active neurons in a color class (Algorithm 2). For these variants detailed balance fails. For example, for a two-spin ferromagnet E=-J s1 s2 with a=e^{-2βJ}, the rejection-free transition probability ratio is P((+,+)->(-,+))/P((- ,+)->(+,+)) = 2a - a^2, whereas the Boltzmann ratio is e^{-2βJ}; these are equal only in the trivial limits. Consequently the stationary distribution is not the Boltzmann distribution of Eq. (1), and the FN/log-cooling convergence theorem of [4] does not transfer to the implemented solvers. Section 3 states that the uncolored version "uses the asynchronous update mechanism given by the noise threshold incorporating the Bernoulli random variable (N^B)" and refers to Algorithm 1, but the pseudocode of SI Algorithm 1 contains no N^B term and no per-step single-spike constraint; Methods 4.6 explicitly says both variants use only the exponential noise N^E. This internal inconsistency is load-bearing, because the abstract's and Section 3's claims of asymptotic ground-state convergence and "reliability" rest on the isomorphism that the implemented algorithms do not satisfy.
  2. For the graph-colored variant, even granting that same-color neurons are conditionally independent, the simultaneous acceptance of multiple flips via independent single-spin threshold tests is not the Metropolis acceptance probability for the joint multi-spin move. Since same-color spins have disjoint clause neighborhoods, the energy change of the joint move is the sum of the individual ΔE_i, but the proposed rule accepts with probability proportional to ∏_i min(1, e^{-ΔE_i/τ}), which is not equal to min(1, e^{-Σ_i ΔE_i/τ}) in general. This violates detailed balance for the joint move as well. No convergence analysis is provided for the colored variant, yet Fig. 2 and Fig. 3 explicitly attribute improved time-to-solution and solution quality to graph coloring. The empirical results therefore come from a heuristic whose stationary distribution, if it exists, is not shown to be the target Boltzmann distribution; the paper's claim that the FN annealer's optimality transfers to the colored machine is unsupported.
  3. The claim that the architecture "maintains resource complexity independent of interaction order" (Abstract) and that "the resource complexity always scales as O(n^2) irrespective of the order of interaction" (Section 3) is not supported by the model or by the manuscript's own MAX-SAT mapping. The number of encoder connections is the number of nonzeros of H, S = Σ_k |C_k|. For the MAX-SAT expansion in SI Section S4, each clause of width p contributes Σ_{ℓ=1}^{p} ℓ C(p,ℓ) = p 2^{p-1} nonzero interactions, so S = M_orig p 2^{p-1}, which grows exponentially in the original clause width p. Thus resource complexity is not independent of interaction order unless one fixes the clause set and per-clause order, which is not the regime described in the abstract or Fig. 1b. The comparison in Fig. 1b is also based on a single quadratization scheme (Chancellor's transformation, Methods 4.1), so the broad statement that quadratization in general leads to exponential overhead is not established. This overstatement affects a headline contribution and should be corrected or carefully qualified.
minor comments (5)
  1. Equation (10) gives only the sign equivalence for ΔE_{i,n}; substituting ΔE = 2Σ into the Metropolis condition Eq. (12) yields Σ_k H_{k,i} J_k T_{k,n-1} < -(τ_n/2) log(u/B+ε), not < -τ_n N^E as written in Eq. (13). The factor of two can be absorbed into the hyperparameters, but as written the derivation is not exact and the claim of functional isomorphism is slightly overstated.
  2. The statement that N^E = log(u/B+ε) "follows an exponential distribution" is imprecise: if u is uniform and ε=0, B=1, then -log u is Exp(1), but the shifted and scaled quantity with general B and ε is not exactly exponential. Since the mean of N^E is later treated as a free hyperparameter (Methods 4.4), this is a presentation issue rather than a technical flaw.
  3. The sentence "the resource complexity always scales as O(n^2) irrespective of the order of interaction" is also internally in tension with the following sentences giving O(n×q) for the encoder and O(m×p) for the decoder; the O(n^2) worst case should be stated as a dense-bound assumption, and the dependence on the number of clauses and their orders should be stated explicitly.
  4. The time axis in Fig. 2b is labeled "a.u." and the inset reports an offset of approximately 5×10^5; the text should state explicitly whether these are CPU iterations, FPGA clock cycles, or wall-clock time, and whether the comparison in Fig. 2c uses the same time units as the dSBM data.
  5. There are several typographical errors, e.g., "T able S1" in SI, "MAX-SA T" in the Section 2.3 heading and Fig. 3 caption, and "the chunk" phrasing in Section 4.6; a careful proofread is needed.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the core spin-flip criterion is an exact algebraic identity, and the FN-annealer convergence cited from prior peer-reviewed work is independent evidence rather than a reduction to the paper's own inputs.

full rationale

The central derivation chain is self-contained. Equation (2) and its Methods derivation (Eqs. 4–10) are exact expansions of the energy change ΔE_i = 2∑_k H_{k,i} J_k T_k, so the encoder update is not a fitted prediction but an algebraic identity. The simulated-annealing acceptance step is transformed without loss into the threshold rule of Eqs. (11)–(16), and the decoder parity update reconstructs the clause flips from the same definition T_k = ∏_i s_i^{H_{k,i}}. No benchmark quantity is defined in terms of the solver's output, and no fitted parameter is renamed as a prediction: the A, Δ, C, and τ_0 values are reported as tuned schedule parameters, not as validated predictions. The asymptotic-convergence claim is imported from the FN-annealer theorem in the authors' prior work [4] and from standard simulated-annealing theory [56]; because [4] is a peer-reviewed, parameter-free derivation with stated assumptions that do not include the higher-order autoencoder target, this self-citation is legitimate independent support and does not make the argument circular. One correctness gap should be flagged without raising the circularity score: Section 3 states that Algorithm 1 uses the Bernoulli variable N^B_{i,n} described in Methods 4.3, while SI Algorithm 1 instead computes mask←(q<μ) with only the annealed exponential threshold and then rejects all but one uniformly chosen accepting spin; the detailed-balance guarantee proven for the asynchronous Bernoulli model therefore does not automatically transfer to the rejection-free or graph-colored implementations. That is an unsupported-claim issue, not an equivalence-of-inputs-and-outputs issue, so the circularity score remains 0.

Assumptions & free parameters 9 free parameters · 5 assumptions · 0 invented entities

The central derivation introduces no new physical entities: the autoencoder is a rewiring of existing toggle neurons, and the FN annealer is a prior device. The load-bearing free parameters are the annealing hyperparameters (tau0, C, delta, A, noise mean, B, epsilon), several of which are tuned per benchmark instance (MAX-CUT A and delta), meaning the reported TTS numbers include fitted constants. The axioms are dominated by the inherited convergence guarantee from [4] and the unproven ergodicity of the parallel colored updates.

free parameters (9)
  • B (SA acceptance scaling hyperparameter) = not specified numerically in main text
    Appears in Eq. (11) and in Algorithms 1-2 as the acceptance scale; folded into the noise statistics, so the mean of N^E is set to -0.083703 (Fig. 3) or -1 (Fig. 4).
  • epsilon (numerical stability constant in N^E = log(u/B+epsilon)) = set by hardware resolution, not quantified
    Defined in Methods 4.3 as a hyperparameter.
  • tau0 (initial temperature scale of the FN schedule) = 0.15625 (Fig. 3 and Fig. 4d), 5e-4 (Fig. 4a-c), or A=tau0*C with A=30/20/30 for G4/G11/G15
    Chosen per benchmark class; MAX-CUT values are picked per graph.
  • C (time constant in tau(t)=tau0/log(1+t/C)) = 8e4 for all experiments
    Stated without derivation; controls annealing speed.
  • delta (sampling period of the discretized FN schedule) = 2e-3 for Fig. 3; best-of-three of {3e-4,3e-5,3e-6} for MAX-CUT; swept 4e-3..4 in Fig. 4
    Annealing speed chosen per instance; the MAX-CUT TTS numbers use the best-performing delta, a form of fitting.
  • mean of N^E (exponential noise variable) = -0.083703 (Fig. 3), -1 (Fig. 4)
    Effectively sets the B/epsilon hyperparameters; tuned for each experiment set.
  • A_asynch (amplitude of the Bernoulli term in mu = tau*N^E + A*N^B) = not specified; only required to be much larger than |tau*log(...)|
    Theoretical asynchrony control; not used in the actual implementations.
  • eta (Bernoulli probability in Eq. (15)) = not specified
    Probability that a neuron is blocked; never quantified.
  • theta (decoder threshold) = not specified
    Gating threshold in Eq. (18) to suppress spurious latent activity; no value given.
assumptions (5)
  • domain assumption The FN-annealer schedule tau(t)=tau0/log(1+t/C) is an optimal annealing schedule guaranteeing asymptotic convergence to the Ising ground state.
    Taken from prior work [4] and not re-derived here; the present paper's convergence claim inherits this guarantee (Methods 4.4).
  • domain assumption The i.i.d. exponential and Bernoulli noise combination preserves irreducibility and aperiodicity of the simulated annealing chain.
    Invoked in Methods 4.3-4.4; the paper states the combination 'ensures that every neuron maintains a nonzero probability of firing'.
  • ad hoc to paper Simultaneous updates of graph-colored, conditionally independent latent neurons do not bias the sampling away from the simulated-annealing stationary distribution.
    Asserted in Methods 4.6 and SI Algorithm 2, but no proof is given; Section 3 concedes that dense graphs require the uncolored asynchronous version to satisfy ergodicity.
  • domain assumption The 16-bit quantized thresholds used on the FPGA inherit the convergence and quality properties of the continuous FN schedule.
    The paper notes (citing [4]) that reduced precision increases time-to-solution but still reports FPGA quality results without a formal quantization analysis (Methods 4.6, Section 2.4).
  • standard math Maximizing the sum of clause polynomials Phi_k(s) in SI S4 is equivalent to the MAX-k-SAT objective up to constants.
    The polynomial construction in SI S4 is exact arithmetic from the CNF definition; no issue.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Higher-Order Neuromorphic Ising Machines -- Autoencoders and Fowler-Nordheim Annealers are all you need for Scalability." pith.science (2026). https://pith.science/paper/G6AEFKBZ

@misc{pith2026250619964,
  author       = {Pith},
  title        = {Pith review of: Higher-Order Neuromorphic Ising Machines -- Autoencoders and Fowler-Nordheim Annealers are all you need for Scalability},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/G6AEFKBZ}},
  note         = {Machine review of arXiv:2506.19964}
}
read the original abstract

We report a higher-order neuromorphic Ising machine that exhibits superior scalability compared to architectures based on quadratization, while also achieving state-of-the-art quality and reliability in solutions with competitive time-to-solution metrics. At the core of the proposed machine is an asynchronous autoencoder architecture that captures higher-order interactions by directly manipulating Ising clauses instead of Ising spins, thereby maintaining resource complexity independent of interaction order. Asymptotic convergence to the Ising ground state is ensured by sampling the autoencoder latent space defined by the spins, based on the annealing dynamics of the Fowler-Nordheim quantum mechanical tunneling. To demonstrate the advantages of the proposed higher-order neuromorphic Ising machine, we systematically solved benchmark combinatorial optimization problems such as MAX-CUT and MAX-SAT, comparing the results to those obtained using a second-order Ising machine employing the same annealing process. Our findings indicate that the proposed architecture consistently provides higher quality solutions in shorter time frames compared to the second-order model across multiple runs. Additionally, we show that the techniques based on the sparsity of the interconnection matrix, such as graph coloring, can be effectively applied to higher-order neuromorphic Ising machines, enhancing the solution quality and the time-to-solution. The time-to-solution can be further improved through hardware co-design, as demonstrated in this paper using a field-programmable gate array (FPGA). The results presented in this paper provide further evidence that autoencoders and Fowler-Nordheim annealers are sufficient to achieve reliability and scaling of any-order neuromorphic Ising machines.

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

63 extracted references · 27 canonical work pages

  1. [4]

    Nature Communications16(1), 3086 (2025) https://doi.org/10.1038/ s41467-025-58231-5

    Chen, Z., Xiao, Z., Akl, M., Leugring, J., Olajide, O., Malik, A., Dennler, N., Harper, C., Bose, S., Gonzalez, H.A., Samaali, M., Liu, G., Eshraghian, J., Pignari, R., Urgese, G., Andreou, A.G., Shankar, S., Mayr, C., Cauwenberghs, G., Chakrabartty, S.: On-off neuromorphic ising machines using fowler-nordheim annealers. Nature Communications16(1), 3086 (...

    work page 2025

  2. [1]

    Science382(6668), 329–335 (2023) https://doi.org/10.1126/ science.adh1174 https://www.science.org/doi/pdf/10.1126/science.adh1174

    Modha, D.S., Akopyan, F., Andreopoulos, A., Appuswamy, R., Arthur, J.V., Cas- sidy, A.S., Datta, P., DeBole, M.V., Esser, S.K., Otero, C.O., Sawada, J., Taba, B., Amir, A., Bablani, D., Carlson, P.J., Flickner, M.D., Gandhasri, R., Garreau, G.J., Ito, M., Klamo, J.L., Kusnitz, J.A., McClatchey, N.J., McKinstry, J.L., Nakamura, Y., Nayak, T.K., Risk, W.P.,...

    2023

  3. [2]

    Science345(6197), 668–673 (2014) https://doi.org/10.1126/science.1254642 https://www.science.org/doi/pdf/10.1126/science.1254642

    Merolla, P.A., Arthur, J.V., Alvarez-Icaza, R., Cassidy, A.S., Sawada, J., Akopyan, F., Jackson, B.L., Imam, N., Guo, C., Nakamura, Y., Brezzo, B., Vo, I., Esser, S.K., Appuswamy, R., Taba, B., Amir, A., Flickner, M.D., Risk, W.P., Manohar, R., Modha, D.S.: A million spiking-neuron integrated circuit with a scalable communication network and interface. Sc...

    doi:10.1126/science.1254642 2014

  4. [3]

    Proceedings of the National Academy of Sciences 113(41), 11441–11446 (2016) https://doi.org/10.1073/pnas.1604850113 https://www.pnas.org/doi/pdf/10.1073/pnas.1604850113 23

    Esser, S.K., Merolla, P.A., Arthur, J.V., Cassidy, A.S., Appuswamy, R., Andreopoulos, A., Berg, D.J., McKinstry, J.L., Melano, T., Barch, D.R., Nolfo, C., Datta, P., Amir, A., Taba, B., Flickner, M.D., Modha, D.S.: Convolutional networks for fast, energy-efficient neuro- morphic computing. Proceedings of the National Academy of Sciences 113(41), 11441–114...

    doi:10.1073/pnas.1604850113 2016

  5. [5]

    Nature 473(7346), 194–198 (2011) https://doi.org/10.1038/nature10012

    Johnson, M.W., Amin, M.H.S., Gildert, S., Lanting, T., Hamze, F., Dickson, N., Harris, R., Berkley, A.J., Johansson, J., Bunyk, P., Chapple, E.M., Enderud, C., Hilton, J.P., Karimi, K., Ladizinsky, E., Ladizinsky, N., Oh, T., Perminov, I., Rich, C., Thom, M.C., Tolkacheva, E., Truncik, C.J.S., Uchaikin, S., Wang, J., Wilson, B., Rose, G.: Quantum annealin...

    doi:10.1038/nature10012 2011

  6. [6]

    Nature617(7959), 61–66 (2023) https: //doi.org/10.1038/s41586-023-05867-2

    King, A.D., Raymond, J., Lanting, T., Harris, R., Zucca, A., Altomare, F., Berkley, A.J., Boothby, K., Ejtemaee, S., Enderud, C., Hoskinson, E., Huang, S., Ladizinsky, E., MacDonald, A.J.R., Marsden, G., Molavi, R., Oh, T., Poulin- Lamarre, G., Reis, M., Rich, C., Sato, Y., Tsai, N., Volkmann, M., Whittaker, J.D., Yao, J., Sandvik, A.W., Amin, M.H.: Quant...

    doi:10.1038/s41586-023-05867-2 2023

  7. [7]

    Nature Photonics8(12), 937–942 (2014) https://doi.org/10.1038/nphoton.2014.249

    Marandi, A., Wang, Z., Takata, K., Byer, R.L., Yamamoto, Y.: Network of time- multiplexed optical parametric oscillators as a coherent ising machine. Nature Photonics8(12), 937–942 (2014) https://doi.org/10.1038/nphoton.2014.249

    doi:10.1038/nphoton.2014.249 2014

  8. [8]

    Science354(6312), 614–617 (2016) https://doi.org/10.1126/science.aah5178 https://www.science.org/doi/pdf/10.1126/science.aah5178

    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., Yamamoto, Y.: A fully pro- grammable 100-spin coherent ising machine with all-to-all connections. Science354(6312), 614–617 (2016) https://doi.org/10.1126/science.aah5178 https://www.s...

    doi:10.1126/science.aah5178 2016

Show all 63 references
  1. [9]

    In: McQuillan, I., Seki, S

    Wang, T., Roychowdhury, J.: Oim: Oscillator-based ising machines for solving combinatorial optimisation problems. In: McQuillan, I., Seki, S. (eds.) Uncon- ventional Computation and Natural Computation, pp. 232–256. Springer, Cham (2019)

    2019

  2. [10]

    Scientific Reports9(1), 14786 (2019) https://doi.org/10

    Chou, J., Bramhavar, S., Ghosh, S., Herzog, W.: Analog coupled oscillator based weighted ising machine. Scientific Reports9(1), 14786 (2019) https://doi.org/10. 1038/s41598-019-49699-5

    2019

  3. [11]

    Scientific Reports7(1), 44370 (2017) https://doi.org/ 10.1038/srep44370

    Sutton, B., Camsari, K.Y., Behin-Aein, B., Datta, S.: Intrinsic optimization using stochastic nanomagnets. Scientific Reports7(1), 44370 (2017) https://doi.org/ 10.1038/srep44370

    2017 doi

  4. [12]

    Nature Electronics5(7), 460–468 (2022) https://doi.org/ 10.1038/s41928-022-00774-2

    Aadit, N.A., Grimaldi, A., Carpentieri, M., Theogarajan, L., Martinis, J.M., Finocchio, G., Camsari, K.Y.: Massively parallel probabilistic computing with sparse ising machines. Nature Electronics5(7), 460–468 (2022) https://doi.org/ 10.1038/s41928-022-00774-2

    2022 doi

  5. [13]

    Proceedings of the National Academy 24 of Sciences79(8), 2554–2558 (1982) https://doi.org/10.1073/pnas.79.8.2554 https://www.pnas.org/doi/pdf/10.1073/pnas.79.8.2554

    Hopfield, J.J.: Neural networks and physical systems with emergent col- lective computational abilities. Proceedings of the National Academy 24 of Sciences79(8), 2554–2558 (1982) https://doi.org/10.1073/pnas.79.8.2554 https://www.pnas.org/doi/pdf/10.1073/pnas.79.8.2554

    1982 doi

  6. [14]

    Nature Electronics3(7), 409–418 (2020) https: //doi.org/10.1038/s41928-020-0436-6

    Cai, F., Kumar, S., Van Vaerenbergh, T., Sheng, X., Liu, R., Li, C., Liu, Z., Foltin, M., Yu, S., Xia, Q., Yang, J.J., Beausoleil, R., Lu, W.D., Strachan, J.P.: Power-efficient combinatorial optimization using intrinsic noise in mem- ristor hopfield neural networks. Nature Ele...

    2020 doi

  7. [15]

    Nature Communications15(1), 3334 (2024) https://doi.org/10.1038/s41467-024-47642-5

    Maher, O., Jim´ enez, M., Delacour, C., Harnack, N., N´ u˜ nez, J., Avedillo, M.J., Linares-Barranco, B., Todri-Sanial, A., Indiveri, G., Karg, S.: A cmos-compatible oscillation-based vo2 ising machine solver. Nature Communications15(1), 3334 (2024) https://doi.org/10.1038/s41...

    2024 doi

  8. [16]

    Nature Communications16(1), 1545 (2025) https://doi.org/10.1038/s41467-025-56739-4

    Yik, J., Berghe, K., Blanken, D., Bouhadjar, Y., Fabre, M., Hueber, P., Ke, W., Khoei, M.A., Kleyko, D., Pacik-Nelson, N., Pierro, A., Stratmann, P., Sun, P.- S.V., Tang, G., Wang, S., Zhou, B., Ahmed, S.H., Vathakkattil Joseph, G., Leto, B., Micheli, A., Mishra, A.K., Lenz, G...

    2025

  9. [17]

    Scientific Reports13(1), 1515 (2023) https://doi.org/10.1038/s41598-023-28217-8

    Mallick, A., Zhao, Z., Bashar, M.K., Alam, S., Islam, M.M., Xiao, Y., Xu, Y., Aziz, A., Narayanan, V., Ni, K., Shukla, N.: Cmos-compatible ising machines built using bistable latches coupled through ferroelectric transistor arrays. Scientific Reports13(1), 1515 (2023) https://...

    2023 doi

  10. [18]

    Scientific Reports13(1), 22858 (2023) https://doi.org/10.1038/s41598-023-49966-6

    Sharma, A., Burns, M., Hahn, A., Huang, M.: Augmenting an electronic ising machine to effectively solve boolean satisfiability. Scientific Reports13(1), 22858 (2023) https://doi.org/10.1038/s41598-023-49966-6

    2023 doi

  11. [19]

    Scientific Reports13(1), 4005 (2023) https://doi.org/10.1038/s41598-023-31155-0

    Albertsson, D.I., Rusu, A.: Highly reconfigurable oscillator-based ising machine through quasiperiodic modulation of coupling strength. Scientific Reports13(1), 4005 (2023) https://doi.org/10.1038/s41598-023-31155-0

    2023 doi

  12. [20]

    Scientific Reports12(1), 981 (2022) https://doi.org/10.1038/s41598-021-04057-2

    Vaidya, J., Surya Kanthi, R.S., Shukla, N.: Creating electronic oscillator-based ising machines without external injection locking. Scientific Reports12(1), 981 (2022) https://doi.org/10.1038/s41598-021-04057-2

    2022 doi

  13. [21]

    Nature Electronics4(7), 502–512 (2021) https: //doi.org/10.1038/s41928-021-00616-7

    Dutta, S., Khanna, A., Assoa, A.S., Paik, H., Schlom, D.G., Toroczkai, Z., Ray- chowdhury, A., Datta, S.: An ising hamiltonian solver based on coupled stochastic 25 phase-transition nano-oscillators. Nature Electronics4(7), 502–512 (2021) https: //doi.org/10.1038/s41928-021-00616-7

    2021 doi

  14. [22]

    IEEE Journal of Solid-State Circuits51(1), 303–309 (2016) https: //doi.org/10.1109/JSSC.2015.2498601

    Yamaoka, M., Yoshimura, C., Hayashi, M., Okuyama, T., Aoki, H., Mizuno, H.: A 20k-spin ising chip to solve combinatorial optimization problems with cmos annealing. IEEE Journal of Solid-State Circuits51(1), 303–309 (2016) https: //doi.org/10.1109/JSSC.2015.2498601

    2016

  15. [23]

    In: 2019 IEEE Inter- national Solid-State Circuits Conference - (ISSCC), pp

    Takemoto, T., Hayashi, M., Yoshimura, C., Yamaoka, M.: 2.6 a 2×30k-spin mul- tichip scalable annealing processor based on a processing-in-memory approach for solving large-scale combinatorial optimization problems. In: 2019 IEEE Inter- national Solid-State Circuits Conference ...

    2019

  16. [24]

    IEEE Micro38(1), 82–99 (2018) https://doi.org/10.1109/MM

    Davies, M., Srinivasa, N., Lin, T.-H., Chinya, G., Cao, Y., Choday, S.H., Dimou, G., Joshi, P., Imam, N., Jain, S., Liao, Y., Lin, C.-K., Lines, A., Liu, R., Math- aikutty, D., McCoy, S., Paul, A., Tse, J., Venkataramanan, G., Weng, Y.-H., Wild, A., Yang, Y., Wang, H.: Loihi: ...

    2018

  17. [25]

    https://arxiv.org/ abs/1911.02385

    Mayr, C., Hoeppner, S., Furber, S.: SpiNNaker 2: A 10 Million Core Processor System for Brain Simulation and Machine Learning (2019). https://arxiv.org/ abs/1911.02385

    2019 arXiv

  18. [26]

    Frontiers in PhysicsV olume 2 - 2014(2014) https://doi.org/10.3389/fphy.2014.00005

    Lucas, A.: Ising formulations of many np problems. Frontiers in PhysicsV olume 2 - 2014(2014) https://doi.org/10.3389/fphy.2014.00005

    2014

  19. [27]

    Discrete Applied Mathe- matics123(1), 155–225 (2002) https://doi.org/10.1016/S0166-218X(01)00341-9

    Boros, E., Hammer, P.L.: Pseudo-boolean optimization. Discrete Applied Mathe- matics123(1), 155–225 (2002) https://doi.org/10.1016/S0166-218X(01)00341-9

    2002 doi

  20. [28]

    In: 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR’05), vol

    Freedman, D., Drineas, P.: Energy minimization via graph cuts: settling what is possible. In: 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR’05), vol. 2, pp. 939–9462 (2005). https://doi.org/10. 1109/CVPR.2005.143

    2005

  21. [29]

    IEEE Transactions on Pattern Analysis and Machine Intelligence33(6), 1234–1249 (2011) https://doi.org/10.1109/TPAMI.2010.91

    Ishikawa, H.: Transformation of general binary mrf minimization to the first-order case. IEEE Transactions on Pattern Analysis and Machine Intelligence33(6), 1234–1249 (2011) https://doi.org/10.1109/TPAMI.2010.91

    2011 doi

  22. [30]

    Nature Communi- cations15(1), 8977 (2024) https://doi.org/10.1038/s41467-024-53270-w

    Nikhar, S., Kannan, S., Aadit, N.A., Chowdhury, S., Camsari, K.Y.: All-to-all reconfigurability with sparse and higher-order ising machines. Nature Communi- cations15(1), 8977 (2024) https://doi.org/10.1038/s41467-024-53270-w

    2024 doi

  23. [31]

    In: 2024 IEEE International Sympo- sium on Circuits and Systems (ISCAS), pp

    Hizzani, M., Heittmann, A., Hutchinson, G., Dobrynin, D., Van Vaerenbergh, T., Bhattacharya, T., Renaudineau, A., Strukov, D., Strachan, J.P.: Memristor- based hardware and algorithms for higher-order hopfield optimization solver outperforming quadratic ising machines. In: 202...

    2024

  24. [32]

    Nature Com- munications14(1), 6033 (2023) https://doi.org/10.1038/s41467-023-41214-9

    Bybee, C., Kleyko, D., Nikonov, D.E., Khosrowshahi, A., Olshausen, B.A., Som- mer, F.T.: Efficient optimization with higher-order ising machines. Nature Com- munications14(1), 6033 (2023) https://doi.org/10.1038/s41467-023-41214-9

    2023 doi

  25. [33]

    Nature Electronics4(3), 208–217 (2021) https://doi.org/10.1038/s41928-021-00546-4

    Tatsumura, K., Yamasaki, M., Goto, H.: Scaling out ising machines using a multi- chip architecture for simulated bifurcation. Nature Electronics4(3), 208–217 (2021) https://doi.org/10.1038/s41928-021-00546-4

    2021 doi

  26. [34]

    Quantum Machine Intelligence7(1), 6 (2025) https://doi.org/10.1007/s42484-024-00231-7

    Wang, H., Liu, Z., Xie, Z., Li, L., Miao, Z., Cui, W., Pan, Y.: Parallel ising annealer via gradient-based hamiltonian monte carlo. Quantum Machine Intelligence7(1), 6 (2025) https://doi.org/10.1007/s42484-024-00231-7

    2025 doi

  27. [35]

    IEEE Journal of Solid-State Circuits56(1), 165–178 (2021) https://doi.org/10

    Yamamoto, K., Kawamura, K., Ando, K., Mertig, N., Takemoto, T., Yamaoka, M., Teramoto, H., Sakai, A., Takamaeda-Yamazaki, S., Motomura, M.: Statica: A 512-spin 0.25m-weight annealing processor with an all-spin-updates-at-once architecture for combinatorial optimization with co...

    2021

  28. [36]

    IEEE Transactions on Communications, 1–1 (2024) https://doi.org/10.1109/TCOMM.2024.3519519

    Nandi, A., Chakrabartty, S., Thakur, C.S.: Margin propagation based xor-sat solvers for decoding of ldpc codes. IEEE Transactions on Communications, 1–1 (2024) https://doi.org/10.1109/TCOMM.2024.3519519

    2024

  29. [37]

    IEEE Transactions on Electron Devices64(3), 1254–1260 (2017) https://doi.org/10.1109/TED.2016

    Zhou, L., Chakrabartty, S.: Self-powered timekeeping and synchronization using fowler–nordheim tunneling-based floating-gate integrators. IEEE Transactions on Electron Devices64(3), 1254–1260 (2017) https://doi.org/10.1109/TED.2016. 2645379

    2017 doi

  30. [38]

    https://web.stanford.edu/∼yyye/yyye/Gset/ (2003)

    Ye, Y.: The Gset Dataset. https://web.stanford.edu/∼yyye/yyye/Gset/ (2003)

    2003

  31. [39]

    Science Advances7(6), 7953 (2021) https://doi.org/10.1126/sciadv.abe7953 https://www.science.org/doi/pdf/10.1126/sciadv.abe7953

    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) https://doi.org/10.1126/sciadv.abe7953 https://www.science.o...

    2021 doi

  32. [40]

    Europhysics Letters90(6), 66002 (2010) https://doi.org/10.1209/0295-5075/90/66002

    Krzakala, F., Zdeborov´ a, L.: Following gibbs states adiabatically —the energy landscape of mean-field glassy systems. Europhysics Letters90(6), 66002 (2010) https://doi.org/10.1209/0295-5075/90/66002

    2010 doi

  33. [41]

    Sci- ence330(6011), 1639–1640 (2010) https://doi.org/10.1126/science.1189804 https://www.science.org/doi/pdf/10.1126/science.1189804

    Ricci-Tersenghi, F.: Being glassy without being hard to solve. Sci- ence330(6011), 1639–1640 (2010) https://doi.org/10.1126/science.1189804 https://www.science.org/doi/pdf/10.1126/science.1189804

    2010 doi

  34. [42]

    Neuron39(6), 991–1004 (2003) https://doi.org/10.1016/j

    Stopfer, M., Jayaraman, V., Laurent, G.: Intensity versus identity coding in an olfactory system. Neuron39(6), 991–1004 (2003) https://doi.org/10.1016/j. neuron.2003.08.011

    2003 doi

  35. [43]

    Sci- ence Advances5(4), 2372 (2019) https://doi.org/10.1126/sciadv.aav2372 https://www.science.org/doi/pdf/10.1126/sciadv.aav2372 27

    Goto, H., Tatsumura, K., Dixon, A.R.: Combinatorial optimization by simulating adiabatic bifurcations in nonlinear hamiltonian systems. Sci- ence Advances5(4), 2372 (2019) https://doi.org/10.1126/sciadv.aav2372 https://www.science.org/doi/pdf/10.1126/sciadv.aav2372 27

    2019 doi

  36. [44]

    Br´ elaz, D.: New methods to color the vertices of a graph. Commun. ACM22(4), 251–256 (1979) https://doi.org/10.1145/359094.359101

    1979

  37. [45]

    283–292 (2000)

    Hoos, H., St¨ utzle, T.: SATLIB: An online resource for research on SAT, pp. 283–292 (2000)

    2000

  38. [46]

    Journal on Satisfiability, Boolean Modeling and Computation11, 133–154 (2019) https://doi.org/10.3233/ SAT190120

    Heule, M., J¨ arvisalo, M., Suda, M.: Sat competition 2018. Journal on Satisfiability, Boolean Modeling and Computation11, 133–154 (2019) https://doi.org/10.3233/ SAT190120

    2019

  39. [47]

    In: Proceedings of the 12th International Joint Conference on Artificial Intelli- gence - Volume 1

    Cheeseman, P., Kanefsky, B., Taylor, W.M.: Where the really hard problems are. In: Proceedings of the 12th International Joint Conference on Artificial Intelli- gence - Volume 1. IJCAI’91, pp. 331–337. Morgan Kaufmann Publishers Inc., San Francisco, CA, USA (1991)

    1991

  40. [48]

    4121, pp

    Prestwich, S., Lynce, I.: Local search for unsatisfiability, vol. 4121, pp. 283–296 (2006). https://doi.org/10.1007/11814948 28

    2006 doi

  41. [49]

    Pierog, T., Karpenko, I., Katzy, J.M., Yatsenko, E., Werner, K.: Epos lhc: Test of collective hadronization with data measured at the cern large hadron collider. Phys. Rev. C92, 034906 (2015) https://doi.org/10.1103/PhysRevC.92.034906

    2015 doi

  42. [50]

    Nature Communications13(1) (2022) https://doi.org/10.1038/s41467-022-29320-6

    Mehta, D., Rahman, M., Aono, K., Chakrabartty, S.: An adaptive synaptic array using fowler–nordheim dynamic analog memory. Nature Communications13(1) (2022) https://doi.org/10.1038/s41467-022-29320-6

    2022 doi

  43. [51]

    Aono A.: A self-powered analog sensor-data-logging device based on fowler-nordheim dynamical systems

    Mehta D., C.S. Aono A.: A self-powered analog sensor-data-logging device based on fowler-nordheim dynamical systems. Nature Communications11(5446) (2020) https://doi.org/10.1038/s41467-020-19292-w

    2020 doi

  44. [52]

    In: 2023 International Electron Devices Meeting (IEDM), pp

    Chen, H.J.-Y., Beauchamp, M., Toprasertpong, K., Huang, F., Le Coeur, L., Nemec, T., Wong, H.-S.P., Boahen, K.: Multi-gate fefet discriminates spatiotem- poral pulse sequences for dendrocentric learning. In: 2023 International Electron Devices Meeting (IEDM), pp. 1–4 (2023). h...

    2023

  45. [53]

    Nature612(7938), 43–50 (2022) https://doi.org/10.1038/s41586-022-05340-6

    Boahen, K.: Dendrocentric learning for synthetic intelligence. Nature612(7938), 43–50 (2022) https://doi.org/10.1038/s41586-022-05340-6

    2022 doi

  46. [54]

    Advances in Mathematics288, 985–1068 (2016) https://doi.org/10.1016/j.aim.2015.11.007

    Coja-Oghlan, A., Panagiotou, K.: The asymptotic k-sat threshold. Advances in Mathematics288, 985–1068 (2016) https://doi.org/10.1016/j.aim.2015.11.007

    2016 doi

  47. [55]

    Scientific Reports6(1), 37107 (2016) https://doi.org/10.1038/srep37107

    Chancellor, N., Zohren, S., Warburton, P.A., Benjamin, S.C., Roberts, S.: A direct mapping of max k-sat and high order parity checks to a chimera graph. Scientific Reports6(1), 37107 (2016) https://doi.org/10.1038/srep37107

    2016 doi

  48. [56]

    Science220(4598), 671–680 (1983) https://doi.org/10.1126/science.220.4598.671

    Kirkpatrick, S., Gelatt, C.D., Vecchi, M.P.: Optimization by simulated annealing. Science220(4598), 671–680 (1983) https://doi.org/10.1126/science.220.4598.671

    1983 doi

  49. [57]

    Frontiers in PhysicsV olume 7 - 2019(2019) https://doi.org/ 10.3389/fphy.2019.00048 28

    Aramon, M., Rosenberg, G., Valiante, E., Miyazawa, T., Tamura, H., Katzgraber, H.G.: Physics-inspired optimization for quadratic unconstrained problems using a digital annealer. Frontiers in PhysicsV olume 7 - 2019(2019) https://doi.org/ 10.3389/fphy.2019.00048 28

    2019

  50. [58]

    Hen, I.: Equation planting: A tool for benchmarking ising machines. Phys. Rev. Appl.12, 011003 (2019) https://doi.org/10.1103/PhysRevApplied.12.011003 29 Supplementary Information for Higher-Order Neuromorphic Ising Machines - Autoencoders and Fowler-Nordheim Annealers are all...

    2019 doi

  51. [59]

    Literal-to-spin mapping: For, each clause Φk, we substitute the spin variables as follows, sk,i = ( +si, ℓ k,i =b i, −si, ℓ k,i =¬b i. (12)

  52. [60]

    Expansion into a higher–order Ising polynomial: Substituting these spin variables intoPM k=1 Φk(s) and collecting like terms yields E(s) =− pX k=1 X 1≤i 1<···<ik≤N J (k) i1i2···ik kY m=1 sim, where each coefficientJ (k) i1···ik arises from collecting multiple spin tuples of th...

  53. [61]

    Each nonzero interaction termJ (k) i1···ik Qk m=1sim corresponds to a row of eHwith ones in columnsi 1,...,i k (and zeros elsewhere), carrying the weightJ (k) i1···ik

    Assembling eH: Finally, we follow the procedure of Section 4.2 to form the binary matrix eH. Each nonzero interaction termJ (k) i1···ik Qk m=1sim corresponds to a row of eHwith ones in columnsi 1,...,i k (and zeros elsewhere), carrying the weightJ (k) i1···ik. S5 Quadratizatio...

  54. [62]

    Literal-to-spin mapping: Replace each clause-local spins k,i by the global spin si via sk,i = ( +si,ifℓ k,i =b i, −si,ifℓ k,i =¬b i, and,s ak =s M+k

  55. [63]

    parallel-trial

    Collect and assemble: Substitute into (13), expand, and group like terms to obtain H(s) = X 1≤i<j≤M+N Qijsisj + M+NX i=1 hisi, so that the final optimization reads max s∈{±1}M+N H(s) = 1 2 sTQs+h T s. S6 Ratio of the size of Interconnection matrix with problem size Fig. S1This...

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.