Day 2 (June 24)

TBA

I present results from http://arxiv.org/abs/2604.13032 a proposal on how to perform quantum annealing in a photonic setting. An advantage of such a setting is that it avoids the intrinsic connectivity issues found in matter based (superconducting circuit or trapped atom) systems. This scheme is based on a single-rail encoding which defines qubit states in terms of the presence or absence of a single photons. Multiple photon occupation is prevented by a Zeno effect which can be implemented using either two-photon absorption or sum-frequency generation. A continuous interpolation between the two can be considered by modulating the loss rate for sum-frequency generation with a lossy pump mode which is initially empty. The blocking of multiple photon occupation can be viewed as a Zeno blockade on a coherent-state driving mode. For interactions I discuss a scheme based on the Hong-Ou-Mandel effect followed by a similar non-linear element to the mode confinement mechanism. I present numerical simulation results which show the performance of this protocol, and show that coherent (i.e. sum-frequency-generation based) Zeno effects tend to be more performant than counterparts based on absorption. After discussing the theoretical basis of the construction and the numerical results I discuss how it could be implemented using time bin encoding, and how schemes could be developed to mitigate errors introduced by photon loss. I will further discuss work which is in progress to develop artificial thermal baths in photonic systems.

http://arxiv.org/abs/2604.13032

D-Wave annealing quantum computers have shown steady increases in complexity. Pushing parametric tolerances to their limits inevitably results in some subset of unyielded components. In this talk I identify complex sub-processor architectures (programmable subgraphs of typical Advantage2 processors) that can act as robust design spaces. These can be used for development of benchmarks, blockchains or for machine learning with robustness to graph or processor changes.

D-Wave’s new analog-digital capabilities open up broad areas of research, allowing detailed investigation of quantum dynamics. But this also brings about the need for more detailed calibration refinement, or “shimming”. In this talk I will walk through the basic steps of calibration refinement using the available per-qubit control knobs, and discuss their effects in cutting-edge quantum simulation experiments.

A. Pérez-Martín1,2, M. Hita-Pérez1, and T. Duty1

1Qilimanjaro Quantum Tech, Carrer de Veneçuela 74, 08019 Barcelona, Spain.

2 Departament de Física Quàntica i Astrofísica, Facultat de Física, Universitat de Barcelona, 08028 Barcelona, Spain.

Superconducting circuits have emerged as a leading platform for quantum computation. Built from macroscopic elements such as inductors, capacitors and Josephson junctions, they offer great flexibility in fabrication and parameter tunability. This versatility enables the realization of a wide range of Hamiltonians, making them particularly suitable for Adiabatic Quantum Computation (AQC). Although some approaches have been proposed [1, 2], current superconducting architectures remain constrained by the limited form of achievable qubit-qubit interactions.

Access to more general interactions would allow, for example, the study of many-body systems [3] and the implementation of non-stoquastic Hamiltonians [4], which could enhance the performance of adiabatic computations and are believed to be a key ingredient for universal AQC [5].

In this work, we analyze a coupling scheme between two fluxonium qubits and derive the effective Hamiltonian governing the system dynamics. The qubits are coupled through a SQUID and a capacitor in parallel. We show that this configuration enables effective fluxonium-fluxonium interactions along the XX, YY and ZZ directions. Our results provide a pathway toward more general Hamiltonian engineering in superconducting platforms and support the development of AQC architectures capable of exploring non-stoquastic regimes.

References

[1] Hita Pérez, M. (2025). Design and modelization of new superconducting circuits for adiabatic quantum computation and other quantum technologies.

[2] Ozfidan, I., Deng, C., Smirnov, A. Y., Lanting, T., Harris, R., Swenson, L., ... & Amin, M. H. (2020). Demonstration of a nonstoquastic Hamiltonian in coupled superconducting flux qubits. Physical Review Applied, 13(3), 034037.

[3] Kounalakis, M., Dickel, C., Bruno, A., Langford, N. K., & Steele, G. A. (2018). Tuneable hopping and nonlinear cross-Kerr interactions in a high-coherence superconducting circuit. npj Quantum Inf. 4.

[4] Klassen, J., & Terhal, B. M. (2019). Two-local qubit Hamiltonians: when are they stoquastic?. Quantum, 3, 139.

[5] Biamonte, J. D., & Love, P. J. (2008). Realizable Hamiltonians for universal adiabatic quantum computers. Physical Review A—Atomic, Molecular, and Optical Physics, 78(1), 012352.

The repetition code has been recruited to preserve information in the presence of noise during adiabatic quantum computation [1]. To suppress thermal excitation of the repetition code, one can use a d-dimensional Ising Hamiltonian; this is effective provided that d > 1 as shown by the Peierls argument [2]. Here, we consider a potentially simpler Ising Hamiltonian in which each qubit is coupled only to a single central qubit. We show that this layout of qubits, although it has just 1 coupling per qubit like the d=1 Ising Hamiltonian, is stable against thermal excitation. This suggests that it could be possible to resist certain forms of noise during adiabatic quantum computation without too many extra qubit-qubit couplers.

[1] K. Pudenz, T. Albash, and D. Lidar, ``Error-corrected quantum annealing with hundreds of qubits,'' Nat. Commun. 5, 3243 (2014).

[2] R. Peierls, ``On Ising's model of ferromagnetism, '' Proc. Cambridge Phil. Soc. 32, 477 (1936); R. B. Griffiths, ``Peierls Proof of Spontaneous Magnetization in a Two-Dimensional Ising Ferromagnetic,'' Phys. Rev. 136, A437 (1964).

Although quantum annealing is usually considered as a method for locating the ground states of difficult spin-glass and optimization problems, its use in approximate optimization --- finding low- but not zero-energy states in a reasonably short amount of time --- is no less important. The traditional (mean-field) theory of spin glasses makes a very underwhelming prediction concerning approximate optimization, that there exists a unique ``threshold'' energy at which all quench and annealing dynamics become trapped until exponential timescales. This would suggest that quantum annealing provides no advantage over classical methods, as it can only return the same threshold energy. However, recent work has shown that this prediction is generically incorrect as far as simulated annealing is concerned: two-stage quenches in which the system first thermalizes at an intermediate temperature can in fact reach states below the naive threshold.
Here we demonstrate that quantum annealing is also capable of exploiting this effect to locate sub-threshold states in $O(1)$ time, using the well-known mixed $p$-spin model as a tractable case study. Furthermore, we find that quantum annealing can be unambiguously better than simulated annealing at doing so: not only can it reach states as far below the naive threshold, but it can do so significantly faster. More precisely, for an annealing schedule taking time $\tau$, the residual energy under quantum annealing decays as $\tau^{-\alpha}$ with an exponent up to twice as large as that of simulated annealing in the cases considered (and further gains are likely possible). Importantly, by deriving and numerically solving closed integro-differential equations that hold in the thermodynamic limit, our results are free from finite-size effects and hold for annealing times that are unambiguously independent of system size.
arXiv:2603.23602