Abstract: Improvements in the control and coherence of superconducting qubits have enabled the development of noisy intermediate-scale quantum (NISQ) processors, and the exploration of problems addressable by NISQ devices that are intractable to classical computation. In this talk, I shall present a brief summary of superconducting qubit technology and discuss results from small-scale demonstrations of algorithms for quantum simulation [1,2] and machine learning [3]. These experiments highlight the detrimental effect of incoherent and readout errors on computations with noisy devices. While this can be remedied, in theory, with quantum error correction, the resources required for experimental implementations are prohibitively large for the near term. In this context, I shall introduce “error mitigation” techniques [4] that enable access to noise-free estimates of expectation values after the application of a short depth quantum circuit, without requiring any additional hardware resources. In particular, I shall demonstrate how our implementation of a zero-noise extrapolation method extends the computational capability of a noisy quantum processor [2].
[1] A. Kandala, et al Nature, 549, 242 (2017)
[2] A. Kandala, et al Nature, 567, 491 (2019)
[3] V. Havlicek, et al Nature, 567, 209 (2019)
[4] K. Temme, et al PRL, 119, 180509 (2017)
