Abstract:We present a benchmarking protocol for universal quantum computers, achieved through the simulation of random dynamical quantum maps. This protocol provides a holistic assessment of system-wide error rates, encapsulating both gate inaccuracies and the errors associated with mid-circuit qubit measurements and resets. By employing random quantum circuits and segmenting mid-circuit qubit measurement and reset in a repeated fashion, we steer the system of qubits to an ensemble of steady-states. These steady-states are described by random Wishart matrices, and align with the steady-state characteristics previously identified in random Lindbladian dynamics, including the universality property. The protocol assesses the resulting ensemble probability distribution measured in the computational basis, effectively avoiding a tomographic reconstruction. Our various numerical simulations demonstrate the relationship between the final distribution and different error sources. Additionally, we implement the protocol on state-of-the-art transmon qubits provided by IBM Quantum, drawing comparisons between empirical results, theoretical expectations, and simulations derived from a fitted noise model of the device.

What problem does this paper attempt to address?

This paper attempts to solve several key problems in quantum computer benchmarking, especially in evaluating the overall system error rate including quantum gates, measurements, and resets. Specifically: 1. **Evaluation of the overall system error rate**: Current benchmarking methods (such as randomized benchmarking) mainly focus on the average gate error rate, but are unable to capture the characteristics of the entire system, especially the errors caused by mid - circuit measurements and resets. This paper proposes a new protocol to provide a comprehensive system - level error rate evaluation by simulating the stochastic dynamical quantum mapping. 2. **Utilization of steady - state characteristics**: By repeatedly applying random quantum circuits and performing measurements and resets in the middle of the circuits, the qubit system reaches a set of steady - state described by a random Wishart matrix. These steady - state characteristics are consistent with those previously identified in random Lindbladian dynamics and are universal. 3. **Avoidance of full tomography reconstruction**: This protocol directly measures the steady - state probability distribution in the computational basis, thereby avoiding the complex full tomography reconstruction process, simplifying the experimental operation and improving the efficiency. 4. **Comparison between theoretical and practical results**: Through experimental verification on superconducting qubits provided by IBM Quantum (such as ibm_tokyo and ibm_osaka), the relationship between theoretical expectations, simulation results, and actual measurements is demonstrated, further proving the effectiveness and practicality of this protocol. ### Summary The main objective of the paper is to develop a benchmarking protocol that can comprehensively evaluate the performance of quantum computers, especially for the overall system error rate including quantum gates, measurements, and resets. By introducing the concept of stochastic dynamical quantum mapping, this protocol can not only evaluate the system - level error rate, but also effectively avoid the complex full tomography reconstruction process, providing new tools and methods for future quantum computing research. ### Formula summary - **Operator and representation of quantum mapping**: \[ E(\rho)=\sum_{i = 1}^{r}K_i\rho K_i^{\dagger} \] where \(r\) is the Kraus rank, and \(K_i\) is the Kraus operator satisfying the trace - preserving condition \(\sum_i K_i^{\dagger}K_i = I\). - **Steady - state eigenvalue decomposition**: \[ \rho_{ss}=U D U^{\dagger}=\sum_i\lambda_i|\psi_i\rangle\langle\psi_i| \] where \(\lambda_i\) is the eigenvalue and \(|\psi_i\rangle\) is the corresponding eigenvector. - **Marchenko - Pastur distribution**: \[ P_{MP}(\lambda)=\frac{1}{2\pi\kappa}\frac{\sqrt{(\lambda_+-\lambda)(\lambda-\lambda_-)}}{\lambda} \] where \(\kappa=\frac{1}{Nr}\), \(N = 2^n\), \(\lambda_{\pm}=\left(1\pm\frac{1}{\sqrt{r}}\right)^2\). These formulas help to understand the mathematical basis and physical meaning of the benchmarking protocol proposed in the paper.

Quantum Benchmarking via Random Dynamical Quantum Maps

Randomized Benchmarking Using Unitary t-Design for Average Fidelity Estimation of Practical Quantum Circuit

Randomized Benchmarking of Quantum Gates

Deterministic Benchmarking of Quantum Gates

Randomized benchmarking in measurement-based quantum computing

Benchmarking bosonic and fermionic dynamics

Benchmarking near-term quantum computers via random circuit sampling

Randomized Benchmarking Protocol for Dynamic Circuits

Scalable Full-Stack Benchmarks for Quantum Computers

Benchmarking digital quantum simulations above hundreds of qubits using quantum critical dynamics

Randomized Benchmarking of Multiqubit Gates.

Preparing random states and benchmarking with many-body quantum chaos

Benchmarking Quantum Simulators using Ergodic Quantum Dynamics

A randomized benchmarking suite for mid-circuit measurements

Measuring error rates of mid-circuit measurements

Comparisons among the Performances of Randomized-framed Benchmarking Protocols under T1, T2 and Coherent Error Models

Defining Standard Strategies for Quantum Benchmarks

Towards a benchmark for quantum computers based on an iterated post-selective protocol

Benchmarking quantum logic operations relative to thresholds for fault tolerance

Benchmarking quantum computers

Randomized benchmarking with random quantum circuits