Abstract:Quantum computers are noisy at present in the absence of error correction and fault tolerance. Interim methods such as error suppression and mitigation find wide applicability. Another method, which is independent of other error suppression and mitigation, and can be applied in conjunction with them to further lower the noise in the system, is circuit cutting. In this paper, we propose a scheduler that finds the optimum schedule for the subcircuits obtained by circuit cutting on the available set of hardware to (i) maximize the overall fidelity, and (ii) ensure that the predefined maximum execution time for each hardware is not exceeded. The fidelity obtained by this method on various benchmark circuits is significantly better than that of the uncut circuit executed on the least noisy device. The average increase in the fidelity obtained by our method are respectively ~12.3% and ~21% for 10-qubit benchmark circuits without and with measurement error mitigation, even when each hardware was allowed the minimum possible execution time. This noise and time optimized distributed scheduler is an initial step towards providing the optimal performance in the current scenario where the users may have limited access to quantum hardware.

What problem does this paper attempt to address?

The paper attempts to address the problem of how to optimize the execution of quantum circuits to improve overall fidelity and ensure that the maximum execution time predefined for each hardware is not exceeded, given the presence of noise in current quantum computers and the lack of error correction and fault tolerance capabilities. Specifically, the authors propose a distributed scheduler that divides a large quantum circuit into multiple sub-circuits through circuit cutting and finds the optimal scheduling scheme on the available hardware to achieve the following goals: 1. **Maximize overall fidelity**: By selecting appropriate hardware and the optimal layout of sub-circuits, reduce the impact of noise on the quantum circuit, thereby improving overall fidelity. 2. **Ensure not to exceed the predefined maximum execution time**: Ensure that the total execution time does not exceed the predefined maximum value when executing sub-circuits on each hardware. ### Background Current quantum computers are limited in their computational capabilities due to the presence of noise. In the absence of error correction and fault tolerance mechanisms, researchers have proposed various methods to suppress and mitigate noise. Additionally, circuit cutting is an effective method to reduce system noise by dividing the circuit into multiple sub-circuits, executing these sub-circuits independently on different hardware, and then restoring the uncut output distribution through classical post-processing. ### Main Contributions The authors designed an Integer Linear Programming (ILP) model aimed at maximizing fidelity while ensuring that the execution time on each hardware does not exceed the predefined maximum value. Through this method, the authors demonstrated significant improvements in fidelity on different benchmark circuits, particularly achieving an average fidelity increase of approximately 12.3% and 21% (without and with measurement error mitigation, respectively) under the condition of allowing the minimum possible execution time. ### Methodology 1. **Circuit Cutting**: Divide a large quantum circuit into multiple sub-circuits, each containing fewer qubits and gates, thereby reducing noise. 2. **Selecting Appropriate Qubits**: Use the mapomatic method to select qubits with less noise for layout, reducing the impact of noise. 3. **Distributed Scheduler**: Design an Integer Linear Programming model to optimize the scheduling of sub-circuits to different hardware, achieving the goals of maximizing fidelity and minimizing execution time. ### Experimental Results The experimental results show that the proposed scheduling method significantly improves fidelity compared to executing the uncut circuit on the least noisy device. Additionally, by reasonably allocating sub-circuits to different hardware, better performance can be achieved within limited execution time. ### Conclusion This study provides an effective method for optimizing performance in the current quantum computing environment, demonstrating the potential to improve quantum computing performance in practical applications through circuit cutting and distributed scheduling.

Distributed Scheduling of Quantum Circuits with Noise and Time Optimization

Resource-aware scheduling of multiple quantum circuits on a hardware device

Optimized Noise Suppression for Quantum Circuits

Quantum circuit scheduler for QPUs usage optimization

Distributed Quantum Computation with Minimum Circuit Execution Time over Quantum Networks

Efficiently improving the performance of noisy quantum computers

Scalable Circuit Cutting and Scheduling in a Resource-constrained and Distributed Quantum System

A Noise Validation for Quantum Circuit Scheduling Through a Service-Oriented Architecture

Experimental Benchmarking of an Automated Deterministic Error-Suppression Workflow for Quantum Algorithms

Noise-Aware Distributed Quantum Approximate Optimization Algorithm on Near-term Quantum Hardware

Multi-Objective Simulated Annealing-Based Quantum Circuit Cutting for Distributed Quantum Computation

Greedy Algorithm Based Circuit Optimization for Near-Term Quantum Simulation

Understanding the Scalability of Circuit Cutting Techniques for Practical Quantum Applications

Near-Term Distributed Quantum Computation using Mean-Field Corrections and Auxiliary Qubits

A framework of partial error correction for intermediate-scale quantum computers

FragQC: An Efficient Quantum Error Reduction Technique using Quantum Circuit Fragmentation

Quantum Divide and Compute: Exploring the Effect of Different Noise Sources

Optimal Partitioning of Quantum Circuits using Gate Cuts and Wire Cuts

Multi-qubit Lattice Surgery Scheduling