Simulating nonlinear optical processes on a superconducting quantum device

TL;DR

This paper presents a method to simulate nonlinear plasma physics problems on superconducting quantum devices by converting wave interactions into Hamiltonian simulations, employing error mitigation and gate decomposition techniques.

Contribution

The authors develop a quantization approach for nonlinear wave interactions and demonstrate its implementation on a superconducting quantum device, addressing native Hamiltonian limitations.

Findings

Successful simulation of nonlinear plasma interactions on two qubits

Error mitigation techniques improve experimental accuracy

Trade-offs in algorithm choices optimize quantum resource use

Abstract

Simulating plasma physics on quantum computers is difficult because most problems of interest are nonlinear, but quantum computers are not naturally suitable for nonlinear operations. In weakly nonlinear regimes, plasma problems can be modeled as wave-wave interactions. In this paper, we develop a quantization approach to convert nonlinear wave-wave interaction problems to Hamiltonian simulation problems. We demonstrate our approach using two qubits on a superconducting device. Unlike a photonic device, a superconducting device does not naturally have the desired interactions in its native Hamiltonian. Nevertheless, Hamiltonian simulations can still be performed by decomposing required unitary operations into native gates. To improve experimental results, we employ a range of error mitigation techniques. Apart from readout error mitigation, we use randomized compilation to transform undiagnosed coherent errors into well-behaved stochastic Pauli channels. Moreover, to compensate for stochastic noise, we rescale exponentially decaying probability amplitudes using rates measured from cycle benchmarking. We carefully consider how different choices of product-formula algorithms affect the overall error and show how a trade-off can be made to best utilize limited quantum resources. This study provides an example of how plasma problems may be solved on near-term quantum computing platforms.