Tools and features index

AQC-Tensor is a Qiskit addon that uses tensor network methods to compile the initial portion of a circuit into a nearly equivalent approximation, with many fewer layers. In addition to simulating the time evolution of a quantum system, this addon can also be useful in these cases:

• A circuit for which tensor-network simulation achieves a great intermediate state

• A good circuit that prepares an approximation to the target state, but with fewer layers when compiled to the target hardware device

Start with the Qiskit circuit library when beginning the process of constructing circuits and workloads. Available features range from standard single- and multi-qubit gates to pre-built ansatze, and also include custom unitary gates.

Optimization mapper is a Qiskit addon that contains the functionality to model optimization problems. For example, you can use it to create models of binary optimization problems to solve with Qiskit.

To get started building quantum workloads to solve optimization problems, refer to the following community repository, which containing guidelines, best practices, and reference implementations for running quantum optimization algorithms. Also listed is a reference QAOA training pipeline.

Qiskit device benchmarking is a respository for code to run various device-level benchmarks through Qiskit. It is not intended to define a benchmark standard, but instead provides code examples to replicate benchmarking metrics that have been reported in literature or elsewhere.

Once your circuits have been created, they then need to be transpiled against the QPU that will execute your workload. The Qiskit SDK has robust tooling for building custom transpilation pipelines that allow you to carefully tune which transformation and analysis passes are executed (and in which sequence). Note that the simplest way to get started is to use the pre-built pass managers.

Dynamic circuits are powerful tools for measuring qubits in the middle of a quantum circuit execution - and based on the outcome of those mid-circuit measurements, you can then perform classical logic operations within the circuit. Some common uses cases include efficient quantum state preparation, long-range entanglement, and sampling of instantaneous quantum polynomial-like circuits.

Dynamical decoupling (DD) is a technique for suppressing errors due to qubit decoherence. It is primarily useful in regions within a quantum circuit in which one or more qubits site idle while other instructions are being executed. Both Sampler and Estimator primitives have built-in support for applying DD. Alternatively, you can use the PadDynamicalDecoupling transpiler pass in Qiskit, or build your own using the new stretch instruction.

This Qiskit addon utilizes a technique to reduce circuit depth by trimming operations from its end at the cost of more operator measurements. However, this increased cost can be lessened by dropping operators with small coefficients, which don't contribute much to the final estimation. You could experiment with coupling this alongside the AQC-Tensor addon for even lower circuit depths.

The noisy estimator analyzer tool (NEAT) analyzes and predicts the performance of your queries when you are gauging the expected performance of estimator-based workloads. It uses Qiskit Aer to simulate the estimation task classically and efficiently, either exactly or in the presence of noise, and can also convert your Primitive Unified Blocs (PUBs) into Cliffordized circuits.

Qiskit Runtime can test workloads while fine-tuning them before submitting them to a QPU. To use this feature, simply specify one of the fake backends from qiskit_ibm_runtime.fake_provider or use one from the Qiskit Aer library. To utilize this feature, we suggest either Cliffordizing your circuit, or scale down the number of qubits in your workload.

The Estimator and Sampler primitives are used to execute your circuits on QPUs. Estimator computes expectation values of observables with respect to states prepared by quantum circuits, and Sampler samples the output register from quantum circuit execution. Both primitives have built-in error suppression support, and Estimator also has built-in error mitigation methods.

Explore the different execution modes - scheduling strategies for execution - when building quantum workloads, to determine which one is most efficient for a specific workload.

Executor allows you to fine-tune error mitigation and other techniques without sacrificing performance, by providing the ingredients to capture design intents on the client-side, and shifting the costly generation of circuit variants to the server-side. With Executor, users will gain a clearer and more composable model for execution workflows, making it easier to experiment with new techniques, reproduce results, and share methods.

Samplomatic is a library for sampling randomizations of your quantum circuits in exactly the way that you specify. It utilizes the Qiskit SDK's Box annotations to specify regions of a circuit that should have similar noise profiles; it can also can group collections of gates to twirl, or use the same noise model from NoiseLearner.

Many error mitigation techniques, such as PEA and PEC, require an accurate noise model. NoiseLearner is a helper program that returns a Pauli-Lindblad noise model for the input circuit. You can then further tune how PEA and PEC are done in your Estimator job, or re-use the same model in an iterative workload.

Pauli twirling, also known as randomized compiling, is an error suppression technique used to convert the effects of unknown types of noise into one that can be characterized (and thus mitigated).

The Matrix-free measurement mitigation (M3) Qiskit addon is used to reduce measurement error by finding corrected measurement probabilities. Use it for problems that benefit from workloads that estimate the expectation value of observables. M3 also works in the context of dynamic circuits.

This error mitigation technique returns an unbiased estimate of the expectation value, at the expense of greater overhead than other techniques such as ZNE. It reproduces the output of the ideal circuit by executing different noisy circuit instances drawn from a random ensemble defined by the linear combination. The Qiskit Runtime Estimator primitive has built-in support for PEC, which you can enable through the Estimator option.

This error mitigation technique first computes the expectation value at different noise levels, then estimates the ideal result by extrapolating the noisy expectation value results to the zero-noise limit. Since this can be done in multiple ways, a number of noise amplification and extrapolation techniques are available.

Gate folding is a noise amplification process that replaces two-qubit gates with equivalent sequences of two-qubit gates and their inverses. This approach is straightforward and simple to use, but can be imperfect.

Probabilistic error amplification (PEA) is a more sophisticated means of amplifying errors for ZNE. It involves running preliminary experiments to learn a twirled noise model of the circuit, and then uses this model to perform a more accurate error amplification.

This error mitigation technique is used to mitigate the effects of measurement errors by using twirled measurements. It can be used alongside many other techniques that mitigate gate errors, such as ZNE and PEC.

This Qiskit addon uses Pauli propagation to characterize Pauli noise affecting different layers of a circuit, by using a learned noise model and propagating its inverse into the target observable to measure.

The shaded lightcones (SLC) Qiskit addon uses Pauli propagation to reduce the number of error terms accounted for in a noise model, according to the specifics of the target observable. It is useful for reducing the sampling overhead when running PEC-based workloads.

With the Sample-based quantum diagonalization (SQD) Qiskit addon, you can implement a post-processing technique to measure the ground state energies of a quantum system. SQD processes samples from a given circuit to project and diagonalize a target Hamiltonian in a subspace spanned by them, essentially "refining" the results obtained from a workload. Use when you want to obtain the eigenvalues and eigenvectors of quantum systems (such as chemical or lattice models).

The Multi-product formulas (MPF) Qiskit addon is used to post-process workloads that simulate the time evolution of a quantum system. The MPF tool will ingest data such as the number of Trotter steps to prepare and solve an associated system of linear equations, which can then be used to refine the expectation-value measurements of a time-evolved state.

The noise models produced by NoiseLearner might not be accurate for Heron backends due to non-Markovian noise on these devices. An internal study has shown this can be alleviated with measurement-based post-selection. This technique adds a series of RX gates followed by a measurement to the circuit, and in post-processing throws away the "bad" bitstrings. The Qiskit addon utilities package contains functionalities to implement this technique, and NoiseLearnerV3 also has built-in support.

Algorithmiq's Tensor-network Error Mitigation (TEM) method is a hybrid quantum-classical algorithm designed for automating and performing noise mitigation. It accomplishes this by constructing a tensor network representing an approximate inverse of the noise affecting a circuit to obtain unbiased estimates of an observable. TEM is a novel error mitigation method based on post-processing with tensor networks, and it provides unbiased error mitigation with the lowest possible shot overhead on the quantum hardware, minimizing the runtime and hence the costs of experiments. It requires exponentially less shots than probabilistic error cancellation (PEC) and significantly less shots than zero noise extrapolation (ZNE).

This Qiskit Function uses AI-powered error suppression and mitigation techniques, and is agnostic to the type of workload being executed.

The QESEM Qiskit Functioon by QEDMA uses a suite of proprietary error mitigation techniques to improve the results of your workload. These techniques include gate optimization, noise-aware transpilation, error suppression, and unbiased error mitigation.

This Qiskit Function by Kipu Quantum solves unconstrained binary optimization problems with the QUBO (Quadratic Unconstrained Binary Optimization) formulation and higher-order (HUBO) optimization problems.

This Qiskit Function by Q-CTRL is designed to solve utility-scale optimization problems. It takes in a high-level problem definition, and executes an entire workflow to optimize the problem, without manual configuration.