Suppressing Correlated Noise: Key to Enhancing Quantum Computing Accuracy

Quantum computing’s potential is hampered by correlated noise, a type of error that can significantly disrupt computation. This paper presents methods and experimental results for suppressing these errors, introducing a Context-Aware Dynamical Decoupling (CADD) framework. This framework can suppress dominant sources of error, making error mitigation or correction less costly. The research suggests that by suppressing correlated noise, quantum computers can perform computations with greater accuracy and less overhead, potentially leading to a significant reduction in sampling overhead for a circuit consisting of a moderate number of layers. This represents a significant step in overcoming challenges associated with scaling up quantum computers.

What is the Impact of Correlated Noise on Quantum Computing?

Quantum computing is a rapidly evolving field that promises to revolutionize the way we process information. However, as quantum computers scale up, it becomes crucial to characterize, suppress, mitigate, and ultimately correct errors in the computation. One of the most detrimental types of errors in quantum computing is correlated noise. This type of noise can occur as a result of spatial and temporal configurations of instructions executing on the quantum processor. Correlated noise, especially those that occur among a set of qubits, can harm computation significantly more than incoherent stochastic errors.

Correlated noise can spread errors faster and make them particularly challenging to handle in error correction protocols. Therefore, accurate characterization and suppression of correlated noise have become an important goal in the field of quantum computing. In this context, the paper presents methods and experimental results for suppressing a wide range of correlated coherent errors using the context of the hardware and the circuits that run on it. The paper also presents a comprehensive characterization of several sources of errors and discusses different strategies that suit each one.

How Can Correlated Noise be Suppressed in Quantum Computing?

The paper introduces a Context-Aware Dynamical Decoupling (CADD) framework that can dress arbitrary circuits with appropriate Dynamical Decoupling (DD) sequences. DD is an effective method to suppress coherent and temporally correlated single-qubit noise in circuits. However, going beyond the single-qubit case and effectively inserting DD in a quantum circuit at scale is a challenging task.

The CADD framework works by coloring the qubits interaction graph at each layer of the circuit based on the contents of that layer and the underlying hardware. This context-aware compiler thus suppresses some dominant sources of error, making further error mitigation or error correction substantially less expensive. For example, the experiments show an increase of 18.5% in layer fidelity for a candidate 10-qubit circuit layer compared to context-unaware suppression.

What is the Role of Error Suppression and Mitigation in Quantum Computing?

Error suppression via better calibration or compilation is often the first line of defense in quantum computing as it can prevent errors from surfacing with only small constant overhead. On the other hand, error mitigation has shown great success in removing errors from large computations but involves additional circuit executions to improve the results, effectively trading exponential sample complexity for accuracy.

In the near term, the intersection of error suppression and error mitigation deserves special attention. Error mitigation techniques typically use additional samples of the circuit and post-processing to improve the accuracy of the results. Even modest improvements in the errors can have a significant impact on the total runtime of an error mitigated computation. Therefore, it is crucial to suppress the known coherent errors in the circuit before twirling and converting those errors to incoherent ones.

How Does the Context-Aware Dynamical Decoupling Framework Work?

The CADD framework is designed to suppress known sources of crosstalk and temporally correlated noise. However, there are cases where it is difficult or undesirable to use DD. This could be the case, for example, when qubits are all executing at the same time, or when the noise spectrum on a qubit is unknown or rapidly changing.

In such cases, the CADD framework can be used to insert appropriate DD sequences into the circuit. The compiler must be aware of different crosstalk terms in the device Hamiltonian, noise spectrums on the qubits, and the physical gate calibrations used. On the other hand, the temporal and spatial structure of the circuit being executed plays an important role in determining the best sequence of DD gates to use.

What are the Implications of this Research for the Future of Quantum Computing?

The research presented in this paper has significant implications for the future of quantum computing. By suppressing correlated noise, quantum computers can perform computations with greater accuracy and less overhead. This could potentially lead to several orders of magnitude reduction of sampling overhead for a circuit consisting of a moderate number of layers.

Furthermore, the CADD framework provides a practical method for suppressing a wide range of correlated coherent errors, which could make quantum computing more reliable and efficient. This research represents a significant step forward in the ongoing effort to overcome the challenges associated with scaling up quantum computers.