Experimental demonstration of boson sampling as a hardware accelerator for monte carlo integration
M. Correa Anguita, T. Roelink, S. Marzban, W.J. Briels, C. Filippi, J. J. Renema · 2025
Read Experimental demonstration of boson sampling as a hardware accelerator for monte carlo integration by M. Correa Anguita, T. Roelink, S. Marzban, W.J. Briels, C. Filippi, J. J. Renema (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-25404.txt.
Introduction
Malaquias Correa Anguita and co-authors test a light-based quantum tool as a boost for hard sums in a 2025 paper. Some sums need vast random trials, and the trials must come from a spread that is hard to fake on a normal machine. The team splits each sum into a spread the quantum tool can sample and a part a normal machine can work out.
They run a proof test on a chip that steers light, working out a small fix to the energy of three trapped particles. The results match theory and computer checks, with gaps traced to known flaws. The team also logs each source of error and its size.
Catalog record
| CREATOR | M. Correa Anguita, T. Roelink, S. Marzban, W.J. Briels, C. Filippi, J. J. Renema |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 56 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-25404
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Experimental demonstration of boson sampling as a hardware accelerator for monte carlo integration M. Correa Anguita, T. Roelink, S. Marzban, W.J. Briels, C. Filippi, J. J. Renema Abstract We present an experimental demonstration of boson sampling as a hardware accelerator for Monte Carlo integration. Our approach leverages importance sampling to factorize an integrand into a distribution that can be sampled using quantum hardware and a function that can be evaluated classically, enabling hybrid quantum-classical computation. We argue that for certain classes of integrals, this method offers a quantum advantage by efficiently sampling from probability distributions that are hard to simulate classically. We also identify structural criteria that must be satisfied to preserve computational hardness, notably the sensitivity of the classical post-processing function to high-order quantum correlations. To validate our protocol, we implement a proof-of-principle experiment on a programmable photonic platform to compute the first-order energy correction of a three-boson system in a harmonic trap under an Efimov-inspired three-body perturbation. The experimental results are consistent with theoretical predictions and numerical simulations, with deviations explained by photon distinguishability, discretization, and unitary imperfections. Additionally, we provide an error budget quantifying the impact of these same sources of noise. Our work establishes a concrete use case for near-term photonic quantum devices and highlights a viable path toward practical quantum advantage in scientific computing. I Introduction Quantum computing has the potential to revolutionize our ability to solve computational problems across a wide variety of application domains, including chemistry banchi_molecular_2020 ; li_hybrid_2024 ; kirsopp_quantum_2022 , cryptography shor_algorithms_1994 ; bernstein_post-quantum_2009 , and machine learning lloyd_quantum_2013 ; schuld_supervised_2018 . However, many of these applications require large-scale, fault-tolerant quantum computers, a development which is still awaiting substantial improvements in both qubit quantity and quality. In the absence of such devices, a central question in the field of quantum computing is whether we can find applications for the quantum computing devices which are currently available. These NISQ (Noisy Intermediate-Scale Quantum) devices are characterized by high noise, an absence of error correction, and system sizes at around a few hundred qubits or qumodes preskill_quantum_2018 . Nevertheless, superconducting and photonic quantum devices have been used for demonstrations of a quantum advantage zhu_quantum_2022 ; madsen_quantum_2022 ; deng_gaussian_2023 ; morvan_phase_2024 , i.e. a demonstration of the ability of a quantum computer to outperform a classical device at a well-defined computational task. However, these demonstrations, while a good illustration of the potential power of a quantum computer, do not solve problems of practical interest. Despite several promising candidates amos_one-shot_2020 ; kahanamoku-meyer_classically_2022 ; aaronson_certified_2023 ; aaronson_verifiable_2024 , it is still an open problem to find an algorithm that can run on near-term hardware that has real-world utility. One of the major blocking issues for finding such an algorithm, particularly in photonic devices, has been that many proposed quantum algorithms have failed to properly take into account the sampling nature of the quantum advantage demonstrations. Quantum advantage demonstrations work by sampling from a probability distribution, whose entries are themselves hard to compute, both classically and using quantum devices, but where these devices can sample efficiently from the distribution, which is strongly believed not to be the case for classical devices. Directly elevating a quantum advantage demonstration into a useful algorithm therefore requires finding a task, where efficiently quantum hardware can sample from a distribution that is classically hard to sample from, and has practical utility. In this work, we experimentally demonstrate such an application—proposed in renema_monte_2024 ; andersen_using_2025 ; andersen_estimating_2025 using photonics (boson sampling) as a hardware accelerator for Monte Carlo integration. The central idea is to use importance sampling to represent a factor in the integral as a multidimensional probability distribution in quantum hardware, which that hardware can then sample from efficiently, with the rest of the integral being computed classically. We provide evidence that a class of integrals exist for which this procedure offers a quantum advantage: first, we show that certain probability distributions can be efficiently programmed on a boson sampler. Next, we identify several classes of integrals which do not offer a quantum advantage. We find cases where a quantum advantage is negated because the integral is classically easy, and cases where it is negated because it is hard even for quantum hardware. We take this as evidence that a sweet spot exists for integrals which offer a quantum advantage. Finally, we show an experimental proof-of-principle demonstration of our algorithm by evaluating the first-order perturbation term using the multi-particle Efimov Hamiltonian on a system of hard-shell bosons in a harmonic trap, using a programmable boson sampler. We compute how these results would change with improvements in the size and quality of our quantum hardware, and we provide an error budget for our experiment. Taking into account known sources of noise—such as partial distinguishability, discretization, and unitary imperfections—the experimental results are quantitatively consistent with both theoretical predictions and numerical simulations, with observed deviations attributable to these noise sources. This establishes the practical feasibility of our protocol, identifies possible experimental improvements, and highlights the potential utility of our approach. II Boson sampling We begin by reviewing boson sampling. Boson sampling is a method for demonstrating a quantum advantage, i.e. the ability of a quantum computer to outperform a classical computer, that is based on quantum optics. In recent years, after the initial proposal by Aaronson and Arkhipov in 2011 aaronson_computational_2010 , boson sampling has transitioned from a theoretical proposal to experimental implementations pushing the bounds of computational limits, both in classical and quantum systems. In a boson sampling experiment, n indistinguishable bosons (most commonly photons) are injected into a linear optical network with m modes, described by a unitary matrix U . The input is specified by an occupation pattern \mu_{\mathrm{in}}=(n_{1},n_{2},\dots,n_{m}) , where n_{j} denotes the number of bosons entering input mode j , with \sum_{j}n_{j}=n . At the output, one records another occupation pattern \mu_{\mathrm{out}}=(n^{\prime}_{1},n^{\prime}_{2},\dots,n^{\prime}_{m}) giving the number of bosons detected per mode. The probability of observing a given output pattern is proportional to the squared modulus of the permanent of a submatrix of U : P(\mu_{\mathrm{out}}|\mu_{\mathrm{in}},U)\propto|\mathrm{Perm}(M)|^{2}, (1) where M is obtained by repeating rows and columns of U according to the occupied input and output modes. The permanent of an n\times n matrix A=(A_{ij}) is defined as \mathrm{Perm}(A)=\sum_{\sigma\in S_{n}}\prod_{i=1}^{n}A_{i,\sigma(i)}, (2) where the sum runs over all permutations \sigma of {1,\dots,n} . Computing such permanents is a #P-hard problem for classical computers valiant_complexity_1979 . As a result, boson sampling became one of the promising approaches to experimentally show quantum advantage. While early experiments were showing the feasibility of boson sampling with small numbers of modes and photons broome_photonic_2013 ; spring_boson_2013 ; tillmann_experimental_2013 , recent work has succeeded in scaling these systems up to demonstrate quantum computational advantage, due to technological advances in photonics and quantum interference control. A significant step towards demonstrating quantum computational advantage using boson sampling was made by Zhong et al in 2019 zhong_quantum_2020 . Their initial experiment showed Gaussian boson sampling with 20 photons in 60 modes, which was later scaled up in Jiuzhang 2.0 and 3.0, reaching up to 255 photons in 144 modes and further advancing the capabilities of the platform zhong_phase-programmable_2021 ; deng_gaussian_2023 . These large-scale demonstrations highlight both the rapid progress of boson sampling experiments and