Speedup in classical simulation of Gaussian boson sampling.

Area of Science:

• Quantum Information Science
• Computational Physics
• Quantum Optics

Background:

• Standard boson sampling uses single photons, while Gaussian boson sampling utilizes squeezed states.
• Demonstrating quantum computational supremacy is a key goal in quantum information science.
• Simulating boson sampling experiments is computationally intensive.

Purpose of the Study:

• To establish a lower bound for achieving quantum computational supremacy in Gaussian boson sampling.
• To propose a more efficient method for simulating Gaussian boson sampling.
• To assess the experimental feasibility of Gaussian boson sampling for quantum supremacy.

Main Methods:

• Numerical analysis of computational costs for Gaussian boson sampling.
• Development of an optimized method for calculating transition probabilities.
• Simulation of Gaussian boson sampling with varying numbers of photons on different computing platforms.

Main Results:

• A lower bound for quantum computational supremacy in Gaussian boson sampling was established.
• The proposed method significantly reduces simulation costs.
• Up to 18, 20, and 30 photons can be simulated on a laptop, workstation, and supercomputer, respectively.
• These simulation capabilities are more accessible than for standard boson sampling.

Conclusions:

• Gaussian boson sampling presents a more experimentally accessible route to demonstrating quantum computational supremacy.
• Reduced simulation costs make larger-scale experiments feasible.
• This approach could accelerate progress in quantum computing research.