Work

Simulation of Linear Optical Interferometers (Extended Abstract)

Year: 2022

Type: conference-paper

Abstract: We propose an algorithm to compute the full output distribution of photons passing through linear optical interferometers. More precisely, given n photons at the input of an m-mode interferometer, the proposed algorithm computes the probabilities of all the $\binom{n + m - 1}{m - 1}$ possible output states in a linear time complexity of $O\left( {n\binom{n + m - 1}{m - 1}} \right)$, outperforming the naive method – computing one permanent for each state – by an exponential factor. The benchmarking has been carried out on a quantum machine learning use-case, to both illustrate the gain and give one possible application. Finally, as storing all the probabilities requires a memory space that is exponential in n and m, we discuss and give strong evidence that memory is far more limiting than the time for the computation of the full distribution.

Source: IEEE International Conference on Quantum Computing and Engineering (QCE)

DOI: 10.1109/qce53715.2022.00080

Authors: Nicolas Heurtel, Shane Mansfield, Jean Senellart, Benoit Valiron

Institutions: Quandela (France), Centre National de la Recherche Scientifique, Institut national de recherche en sciences et technologies du numérique, Université Paris-Saclay, CentraleSupélec

Language: English

Topic: Quantum Computing Algorithms and Architecture

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: green

Funder: Agence Nationale de la Recherche

Awards: Etude de la pile quantique : Algorithmes, modèles de calcul et simulation pour l’informatique quantique, Software-based quantum program and resource optimisation