Heavy tails and pruning in programmable photonic circuits

TL;DR

This paper uncovers heavy-tailed distributions in large-scale programmable photonic circuits and demonstrates that strategic pruning of superfluous rotations can enhance fidelity and energy efficiency in quantum and deep learning applications.

Contribution

It reveals the heavy-tailed nature of rotation operators in photonic circuits and introduces a pruning method to improve performance and efficiency.

Findings

Heavy-tailed distributions of rotation operators identified.

Pruning hub phase shifters improves fidelity.

Universal pruning architecture proven effective.

Abstract

Developing hardware for high-dimensional unitary operators plays a vital role in implementing quantum computations and deep learning accelerations. Programmable photonic circuits are singularly promising candidates for universal unitaries owing to intrinsic unitarity, ultrafast tunability, and energy efficiency of photonic platforms. Nonetheless, when the scale of a photonic circuit increases, the effects of noise on the fidelity of quantum operators and deep learning weight matrices become more severe. Here we demonstrate a nontrivial stochastic nature of large-scale programmable photonic circuits-heavy-tailed distributions of rotation operators-that enables the development of high-fidelity universal unitaries through designed pruning of superfluous rotations. The power law and the Pareto principle for the conventional architecture of programmable photonic circuits are revealed with the presence of hub phase shifters, allowing for the application of network pruning to the design of photonic hardware. We extract a universal architecture for pruning random unitary matrices and prove that "the bad is sometimes better to be removed" to achieve high fidelity and energy efficiency. This result lowers the hurdle for high fidelity in large-scale quantum computing and photonic deep learning accelerators.