arXiv:2509.11861

Optimizing Quantum Photonic Integrated Circuits using Differentiable Tensor Networks

Mathias Van Regemortel, Thomas Van Vaerenbergh

Published Sep 15, 2025• 0 citations

Abstract

Recent reports of large photonic nonlinearities in integrated photonic devices, using the strong excitonic light-matter coupling in semiconductors, necessitate a tailored design framework for quantum processing in the limit of low photon occupation. We present a gradient-based optimization method for quantum photonic integrated circuits, which are composed of nonlinear unitary coupling gates and stochastic, nonunitary components for sampling the photonic losses. As core of our method, differentiable tensor-networks are leveraged, which are accurate in the regime of low photonic occupation and modest intermode entanglement. After characterizing the circuit gate architecture with field simulations of GaAs-based samples, we demonstrate the applicability of our method by optimizing quantum photonic circuits for two key use cases: integrated designs for quantum optical state preparation and tailored optimal readout for quantum phase sensing.

DOI: 10.48550/arxiv.2509.11861

Related papers

Variational learning of integrated quantum photonic circuits

Hui Zhang, Chengran Yang, Wai‐Keong Mok +11 more

Integrated photonic circuits play a crucial role in implementing quantum information processing in the noisy intermediate-scale quantum (NISQ) era. Variational learning is a promising avenue that leverages classical optimization techniques to enhance quantum advantages on NISQ devices. However, most variational algorithms are circuit-model-based and encounter challenges when implemented on integrated photonic circuits, because they involve explicit decomposition of large quantum circuits into sequences of basic entangled gates, leading to an exponential decay of success probability due to the non-deterministic nature of photonic entangling gates. Here, we present a variational learning approach for designing quantum photonic circuits, which directly incorporates post-selection and elementary photonic elements into the training process. The complicated circuit is treated as a single nonlinear logical operator, and a unified design is discovered for it through variational learning. Engineering an integrated photonic chip with automated control, we adjust and optimize the internal parameters of the chip in real time for task-specific cost functions. We utilize a simple case of designing photonic circuits for a single ancilla CNOT gate with improved success rate to illustrate how our proposed approach works, and then apply the approach in the first demonstration of quantum stochastic simulation using integrated photonics.

arXiv:2411.12417Nov 19, 20240 citations

Machine learning method for state preparation and gate synthesis on photonic quantum computers

Juan Miguel Arrazola, Thomas R Bromley, Josh Izaac +3 more

Abstract We show how techniques from machine learning and optimization can be used to find circuits of photonic quantum computers that perform a desired transformation between input and output states. In the simplest case of a single input state, our method discovers circuits for preparing a desired quantum state. In the more general case of several input and output relations, our method obtains circuits that reproduce the action of a target unitary transformation. We use a continuous-variable quantum neural network as the circuit architecture. The network is composed of several layers of optical gates with variable parameters that are optimized by applying automatic differentiation using the TensorFlow backend of the Strawberry Fields photonic quantum computer simulator. We demonstrate the power and versatility of our methods by learning how to use short-depth circuits to synthesize single photons, Gottesman–Kitaev–Preskill states, NOON states, cubic phase gates, random unitaries, cross-Kerr interactions, as well as several other states and gates. We routinely obtain high fidelities above 99% using short-depth circuits, typically consisting of a few hundred gates. The circuits are obtained automatically by simply specifying the target state or gate and running the optimization algorithm.

arXiv:1807.10781Quantum Science and TechnologyDec 3, 2018151 citations

Optimization of non-Gaussian state generation using tensor networks and automatic differentiation

Ryutaro Nagai, Takao Tomono

Tensor networks have been eagerly developed in recent years as a simulation method for quantum computation on classical computers. The efficiency of tensor networks for simulating quantum computation has been shown by solving a task designed for demonstrating quantum computing supremacy. However, such a task is not useful for real application. Therefore, there is the room for research of finding specific applications suitable for exploiting tensor network simulation. As such an application, we propose optimization of circuits to generate non- Gaussian states. Non-Gaussian state is one of the crucial elements for achieving universal continuous variable quantum computation. We consider this problem as a variational optimization of optical circuits to generate desired non-Gaussian states. Tensor networks efficiently simulate the optical circuits, and their compatibility with automatic differentiation provides a heuristic way to optimize the circuits. We propose and report the result of our approach that combines tensor networks and automatic differentiation to perform variational optimization of circuits such that the desired non-Gaussian state is approximately generated. The complexity of contracting tensor networks is scaled by degree of entanglement, so we consider relatively shallow optical circuit. Such a restriction is beneficial for considering devices without error correction method, which are accessible in near-term.

IEEE International Conference on Quantum Computing and Engineering (QCE)Sep 1, 20221 citation

Quantum Circuit Optimization using Differentiable Programming of Tensor Network States

David Rogerson, Ananda Roy

Efficient quantum circuit optimization schemes are central to quantum simulation of strongly interacting quantum many body systems. Here, we present an optimization algorithm which combines machine learning techniques and tensor network methods. The said algorithm runs on classical hardware and finds shallow, accurate quantum circuits by minimizing scalar cost functions. The gradients relevant for the optimization process are computed using the reverse mode automatic differentiation technique implemented on top of the time-evolved block decimation algorithm for matrix product states. A variation of the ADAM optimizer is utilized to perform a gradient descent on the manifolds of charge conserving unitary operators to find the optimal quantum circuit. The efficacy of this approach is demonstrated by finding the ground states of spin chain Hamiltonians for the Ising, three-state Potts and the massive Schwinger models for system sizes up to L=100. The first ten excited states of these models are also obtained for system sizes L=24. All circuits achieve high state fidelities within reasonable CPU time and modest memory requirements.

arXiv:2408.12583Aug 22, 20240 citations

Source to detector simulation of quantum photonic integrated circuits

Sebastian Gitt, Jeff F. Young, Dylan McGuire

Quantum photonic devices are typically based on either discrete variable (DV) or continuous variable (CV) principles. In either case, their performance must meet strict “fidelity” requirements, especially when scaled using integrated photonic circuitry. DV devices must be robust to imperfections such as timing delays between photons, unwanted spectral correlations present in photons from nonideal single photon sources and variations in circuit components. Furthermore, it becomes increasingly important to have a methodology for characterizing the performance of devices in the presence of such errors. Here, we present a method for simulating quantum photonic integrated circuits (PICs) and use it to model the behavior of a circuit constructed to model the evolution of a quantum state subject to a Bose-Hubbard Hamiltonian, introducing variations to the coupling gaps in the directional couplers inside the circuit. Input states consisting of both indistinguishable and distinguishable photons are modelled. Compact models for the directional couplers and other photonic circuit elements that are parameterized for a specific fabrication process can be incorporated into the circuit using a foundry-associated compact model library, ensuring consistency between circuit design and manufactured component. In CV applications, the degree of squeezing provided by some nonlinear element is often a key metric. Here we describe how the degree of squeezing produced by spontaneous four-wave mixing (SFWM) in a microring resonator can be modelled in the low-power limit, accounting for the effects of self-phase modulation (SPM), cross-phase modulation (XPM), and component losses.

Smart Photonic and Optoelectronic Integrated CircuitsMar 8, 20240 citations

Towards deterministic quantum photonic information processing with quantum dot phase shifters

Jacob Ewaniuk, Adam McCaw, Sofia Arranz Regidor +3 more

Semiconductor quantum dots (QDs) are a type of solid-state quantum emitter that can act as a near-ideal quantum light-matter interface when integrated with high-quality nanophotonic systems. Though QDs have typically been used to create state-of-the-art, on-demand single photon sources, here we widen the perspective on QDs, showing how to design quantum photonic integrated circuits based on both linear and nonlinear QD phase shifters. Specifically, we find that linear QD phase shifters can be used to realize cryogenically-compatible, fast, low-loss, and high-fidelity reconfigurable linear circuits. When paired with QDs that mediate interactions between photonic qubits, generating nonlinear phase shifts, deterministic quantum photonic logic gates can be achieved. Thus, our work paves the way for the realization of on-chip, cryogenically-compatible linear and nonlinear quantum photonic circuits, including quantum photonic neural networks, which can form the foundation for scalable and efficient quantum photonic technologies.

Photonics for Quantum (PfQ)Aug 2, 20240 citations