Classical Modeling of a Lossy Gaussian Bosonic Sampler.

Area of Science:

• Quantum Information Science
• Computational Physics
• Quantum Optics

Background:

• Gaussian boson sampling (GBS) is a prominent problem for demonstrating quantum advantage.
• Simulating GBS classically is computationally challenging, especially for lossy systems.

Purpose of the Study:

• To develop an efficient classical algorithm for approximating the output distribution of lossy GBS instances.
• To identify conditions that optimize the performance of the proposed classical simulation algorithm.

Main Methods:

• The algorithm employs a Taylor series expansion to approximate the GBS output.
• Computational complexity is polynomial in the number of modes for a fixed number of expansion terms.
• Analysis focuses on the convergence rate of the Taylor series based on input squeezing and loss parameters.

Main Results:

• The proposed algorithm provides an approximate classical simulation of lossy GBS.
• Accuracy increases with the number of terms in the Taylor series expansion.
• Optimal efficiency is achieved under specific conditions of input squeezing and loss levels.

Conclusions:

• The developed algorithm can efficiently simulate recent quantum advantage experiments in GBS.
• The findings provide a benchmark for assessing quantum computational advantage in photonic systems.