Hardware Efficient Quantum Kernels Using Multimode Bulk Acoustic Resonators

TL;DR

This paper proposes a quantum kernel method using multimode bulk acoustic resonators and Kerr nonlinear devices, demonstrating non-classical behavior and potential quantum advantage in kernel computation scalability.

Contribution

It introduces a quantum kernel design utilizing Kerr nonlinearities and acoustic resonators, extending prior work with time-dependent simulations for scalable quantum machine learning.

Findings

Kerr nonlinearity induces non-classical states in multimode systems

Quantum kernel shows potential for computational advantage over classical methods

Scaling analysis indicates classical intractability as resonator number increases

Abstract

The kernel trick is a widely applicable technique in machine learning domains that maps datasets that are difficult to classify into a computationally friendly feature space. As the dimension of the dataset scales, these kernel calculations can quickly become computationally intractable or data inefficient. In this work, we extend prior efforts in quantum kernel design for Kerr nonlinear devices by implementing time-dependent simulations of a Kerr-qubit coupled to acoustic resonators. For experimentally feasible parameters, we demonstrate that the Kerr nonlinearity directly induces non-classical behavior in the multimode system, which we use to define and analyze a quantum-enhanced kernel. Finally, we present a brief scaling characterization that demonstrates the computational intractability of classically simulating the kernel as the number of resonators scales.