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