Understanding the Scalability of Circuit Cutting Techniques for Practical Quantum Applications

Songqinghao Yang,Prakash Murali

2024-11-26

Abstract:Circuit cutting allows quantum circuits larger than the available hardware to be executed. Cutting techniques split circuits into smaller subcircuits, run them on the hardware, and recombine results through classical post-processing. Circuit cutting techniques have been extensively researched over the last five years and it been adopted by major quantum hardware vendors as part of their scaling roadmaps. We examine whether current circuit cutting techniques are practical for orchestrating executions on fault-tolerant quantum computers. We conduct a resource estimation-based benchmarking of important quantum applications and different types of circuit cutting techniques. Our applications include practically relevant algorithms, such as Hamiltonian simulation, kernels such as quantum Fourier transform and more. To cut these applications, we use IBM's Qiskit cutting tool. We estimate resources for subcircuits using Microsoft's Azure Quantum Resource Estimator and develop models to determine the qubit, quantum and classical runtime needs of circuit cutting. We demonstrate that while circuit cutting works for small-scale systems, the exponential growth of the quantum runtime and the classical post-processing overhead as the qubit count increases renders it impractical for larger quantum systems with current implementation strategies. As we transition from noisy quantum hardware to fault-tolerance, our work provides important guidance for the design of quantum software and runtime systems.

Quantum Physics