Noise-Induced Barren Plateaus in Variational Quantum Algorithms

Samson Wang,Enrico Fontana,M. Cerezo,Kunal Sharma,Akira Sone,Lukasz Cincio,Patrick J. Coles

DOI: https://doi.org/10.1038/s41467-021-27045-6

2024-03-02

Abstract:Variational Quantum Algorithms (VQAs) may be a path to quantum advantage on Noisy Intermediate-Scale Quantum (NISQ) computers. A natural question is whether noise on NISQ devices places fundamental limitations on VQA performance. We rigorously prove a serious limitation for noisy VQAs, in that the noise causes the training landscape to have a barren plateau (i.e., vanishing gradient). Specifically, for the local Pauli noise considered, we prove that the gradient vanishes exponentially in the number of qubits $n$ if the depth of the ansatz grows linearly with $n$. These noise-induced barren plateaus (NIBPs) are conceptually different from noise-free barren plateaus, which are linked to random parameter initialization. Our result is formulated for a generic ansatz that includes as special cases the Quantum Alternating Operator Ansatz and the Unitary Coupled Cluster Ansatz, among others. For the former, our numerical heuristics demonstrate the NIBP phenomenon for a realistic hardware noise model.

Quantum Physics,Machine Learning

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: the performance limitations of variational quantum algorithms (VQAs) in noisy environments, especially whether noise will lead to the vanishing - gradient problem (the so - called "barren plateau" phenomenon), thus affecting the training and optimization effects of VQAs.

Specifically, the authors are concerned with whether VQAs running on noisy intermediate - scale quantum (NISQ) computers will encounter fundamental performance bottlenecks due to the presence of hardware noise. They have shown through rigorous mathematical proofs that in the presence of local Pauli noise, as the number of qubits \(n\) increases and the quantum circuit depth \(L\) grows linearly, the gradient will decay exponentially to zero. These noise - induced barren plateaus (NIBPs) are different from the barren plateaus in the noiseless case, where the latter are usually related to the parameter initialization method.

In addition, the paper also explores the impact of this phenomenon on specific applications, such as the quantum approximate optimization algorithm (QAOA) in combinatorial optimization problems and the variational quantum eigensolver (VQE) in chemical calculations. The authors not only provide theoretical analysis but also conduct numerical simulation experiments to verify their conclusions. These research results are of great significance for understanding the limitations of current quantum computing technology in practical applications and provide guidance for improving the design of VQAs in the future.