Session A13: Quantum Characterization, Verification and Validation

Estimating the coherence of noise

To harness the advantages of quantum information processing, quantum systems have to be controlled to within some maximum threshold error. Certifying whether the error is below the threshold is possible by performing full quantum process tomography, however, quantum process tomography is inefficient in the number of qubits and is sensitive to state-preparation and measurement errors (SPAM). Randomized benchmarking has been developed as an efficient method for estimating the average infidelity of noise to the identity. However, the worst-case error, as quantified by the diamond distance from the identity, can be more relevant to determining whether an experimental implementation is at the threshold for fault-tolerant quantum computation. The best possible bound on the worst-case error (without further assumptions on the noise) scales as the square root of the infidelity and can be orders of magnitude greater than the reported average error. We define a new quantification of the coherence of a general noise channel, the unitarity, and show that it can be estimated using an efficient protocol that is robust to SPAM. Furthermore, we also show how the unitarity can be used with the infidelity obtained from randomized benchmarking to obtain improved estimates of the diamond distance and to efficiently determine whether experimental noise is close to stochastic Pauli noise.   

Gate-set tomography and beyond

Four years ago, there was no reliable way to characterize and debug quantum gates.  Process tomography required perfectly pre-calibrated gates, while randomized benchmarking only yielded an overall error rate.  Gate-set tomography (GST) emerged around 2012-13 in several variants (most notably at IBM; see PRA 87, 062119) to address this need, providing complete and calibration-free characterization of gates.  At Sandia, we have pushed the capabilities of GST well beyond these initial goals.  In this talk, I’ll demonstrate our open web interface, show how we characterize gates with accuracy at the Heisenberg limit, discuss how we put error bars on the results, and present experimental GST estimates with 1e-5 error bars.  I’ll also present preliminary results of GST on 2-qubit gates, including a brief survey of the tricks we use to make it possible. I’ll conclude with an analysis of GST’s limitations (e.g., it scales poorly), and the techniques under development for characterizing and debugging larger (3+ qubit) systems.   

Toward a new culture in verified quantum operations

Measuring error rates of quantum operations has become an indispensable component in any aspiring platform for quantum computation. As the quality of controlled quantum operations increases, the demands on the accuracy and precision with which we measure these error rates also grows. However, well-meaning scientists that report these error measures are faced with a sea of non-standardized methodologies and are often asked during publication for only coarse information about how their estimates were obtained. Moreover, there are serious incentives to use methodologies and measures that will continually produce numbers that improve with time to show progress. These problems will only get exacerbated as our typical error rates go from 1 in 100 to 1 in 1000 or less. This talk will survey existing challenges presented by the current paradigm and offer some suggestions for solutions than can help us move toward fair and standardized methods for error metrology in quantum computing experiments, and towards a culture that values full disclose of methodologies and higher standards for data analysis.   

Applying QCVV protocols to real physical systems.

As experimental systems move closer to realizing small-scale quantum computers with high fidelity operations, errors become harder to detect and diagnose. Verification and validation protocols are becoming increasingly important for detecting and understanding the precise nature of these errors. I will outline various methods and protocols currently used to deal with errors in experimental systems. I will also discuss recent advances in implementing high fidelity operations which will help to understand some of the tools that are still needed on the road to realizing larger scale quantum systems.   

Applying benchmarking protocols to encoded qubits with non-Markovian errors

An essential goal for any quantum information processing platform is to develop the tools necessary to validate high-fidelity quantum gates. This effort has produced a suite of benchmarking and tomographic protocols that have been applied to a wide variety of physical implementations. All these protocols, however, were designed with strict error assumptions that can and will be violated by physical errors, especially as we push to lower and lower error rates. In this talk we look at randomized benchmarking with encoded states (from which leakage errors may occur) in the presence of non-Markovian noise and under the influence of sequence-length dependent filtering errors. These circumstances may apply to a variety of physical systems, but are particularly pertinent for 1/f charge noise and hyperfine leakage noise in electrically controlled quantum dot qubits. We demonstrate how these errors affect the outcome of randomized benchmarking, including the signatures of said errors and the confidence with which we can report an average gate fidelity.