انتقل إلى المحتوى الرئيسي

لم تُترجم بعد

هذه الصفحة لم تُترجم بعد. يتم عرض المحتوى باللغة الإنجليزية.

Overview of noise management techniques

When executing quantum workloads, there are multiple ways to reduce the impact of noise. The open-source Qiskit addons provide error mitigation and suppression techniques that integrate directly into your development workflow, while Qiskit Runtime applies advanced error mitigation strategies automatically when jobs are submitted for execution. This page indexes all available tools and features across both options to help you choose the right approach to manage noise when building quantum workloads.

General noise management techniques

Directed execution model

Fine-tune error mitigation and other techniques by capturing design intents on the client side and shifting, and shifting the costly generation of circuit variants to the server side.

Browse documentation →

Dynamical decoupling

Inserts pulse sequences on idling qubits to suppress coherence errors caused by unwanted interactions between qubits during circuit execution.

Noise-tailoring technique that transforms any noise channel into a Pauli channel that has a more specific structure; often combined with other error mitigation techniques that work well with Pauli noise.

AQC-Tensor Qiskit addon

Users can compile the initial portion of a circuit into a nearly equivalent approximation of that circuit, but with fewer layers.

Browse documentation →

Error mitigation for expectation value estimation

Twirled readout error extinction (TREX)

Error mitigation tool within Qiskit Runtime which mitigate the effects of measurement errors by randomly substituting them with a twirled measurement sequence.

Zero-noise extrapolation (ZNE)

Error mitigation technique that computes the expectation value at different noise levels, and then estimates the ideal result by extrapolating the noisy expectation value results to the zero-noise limit.

Probabilistic error amplification (PEA)

ZNE technique that involves running preliminary experiments to learn a twirled noise model of the circuit, then uses this model to perform a more accurate error amplification.

Probabilistic error cancellation (PEC)

Returns an unbiased estimate of the expectation value, at the expense of greater overhead than other techniques such as ZNE. It extrapolates output of the ideal circuit by executing different noisy circuit instances.

PEC with shaded lightcones

A modified PEC technique that uses Pauli propagation to reduce the number of error terms accounted for in a noise model according to the specifics of the target observable.

Browse documentation →

Operator backpropagation (OBP)

Uses a method based on Clifford perturbation theory to reduce circuit depth by trimming operations from its end at the cost of more operator measurements.

Browse documentation →

Propagated noise absorption (PNA)

Technique for mitigating errors in observable expectation values by 'absorbing' the inverses of learned noise channels into an observable using Pauli propagation.

Browse documentation →

Error mitigation for sampling results

Sample-based quantum diagonalization (SQD)

Implements a technique for finding eigenvalues and eigenvectors of quantum operators, such as a quantum system Hamiltonian, using quantum and distributed classical computing together.

Browse documentation →

An HPC-ready implementation of the SQD addon. It is written in modern C++17 standards and is designed to create a single compiled binary for use with MPI.

Browse documentation →

Matrix-free Measurement Mitigation

Matrix-free Measurement Mitigation (M3) is a package for scalable quantum measurement error mitigation that can be computed in parallel.

Browse documentation →

General noise management techniques
Error mitigation for expectation value estimation
Error mitigation for sampling results