Optimal Partitioning of Quantum Circuits using Gate Cuts and Wire Cuts

TL;DR

This paper introduces an optimal quantum circuit partitioning method that minimizes execution overhead by combining gate and wire cuts, significantly reducing resource requirements for near-term quantum computers.

Contribution

The paper presents a novel partitioning approach that optimally combines gate cuts and wire cuts with classical communication, improving efficiency over previous methods.

Findings

Reduces quantum computing time overhead by 73% on average.

Decreases qubit requirements by 40% within a one-hour runtime budget.

Enhances the computational reach of near-term quantum devices.

Abstract

A limited number of qubits, high error rates, and limited qubit connectivity are major challenges for effective near-term quantum computations. Quantum circuit partitioning divides a quantum computation into a set of computations that include smaller-scale quantum (sub)circuits and classical postprocessing steps. These quantum subcircuits require fewer qubits, incur a smaller effort for satisfying qubit connectivity requirements, and typically incur less error. Thus, quantum circuit partitioning has the potential to enable quantum computations that would otherwise only be available on more matured hardware. However, partitioning quantum circuits generally incurs an exponential increase in quantum computing runtime by repeatedly executing quantum subcircuits. Previous work results in non-optimal subcircuit executions hereby limiting the scope of quantum circuit partitioning. In this work, we develop an optimal partitioning method based on recent advances in quantum circuit knitting. By considering wire cuts and gate cuts in conjunction with ancilla qubit insertions and classical communication, the developed method can determine a minimal cost quantum circuit partitioning. Compared to previous work, we demonstrate the developed method to reduce the overhead in quantum computing time by 73% on average for 56% of evaluated quantum circuits. Given a one hour runtime budget on a typical near-term quantum computer, the developed method could reduce the qubit requirement of the evaluated quantum circuits by 40% on average. These results highlight the ability of the developed method to extend the computational reach of near-term quantum computers by reducing the qubit requirement at a lower increase in quantum circuit executions.