Triangle Quantum Computing Seminar Series: Efficient simulation and learning of Markovian noise

ABSTRACT: Classical simulation of noisy quantum circuits is essential for understanding quantum computing experiments. It enables scalable error characterization, analysis of how noise impacts quantum algorithms, and optimized implementations of quantum error correction. However, most existing efficient simulation techniques can only simulate the effects of stochastic noise. The lack of efficient ways to simulate coherent errors, which are common and significant, has frustrated research. In this talk, I'll present an efficient algorithm for approximate simulation of Clifford circuits with arbitrary small Markovian errors (including coherent errors), and how this algorithm can be used to study the impact of coherent errors on syndrome extract circuits for the surface code. I'll then show how this algorithm can be used for efficient characterization of errors in many-qubit systems, using an approximate linearized tomography method, and how this can be combined with neural network methods to potentially enable efficient and easy characterization of both Markovian and non-Markovian errors at scale. BIO: Timothy Proctor is a staff scientist in Sandia's Quantum Performance Laboratory, where he has been working on quantum computer benchmarking and noise modelling for the last 10 years. He obtained his PhD in 2016 from the University of Leeds in the UK, advised by Viv Kendon, where he worked on quantum sensor networks and quantum computing theory. His current interests include scalable benchmarking of quantum computers, modelling noise in quantum error correction, and developing AI/ML-aided methods for efficient noise modelling and characterization. -- Co-hosted by the Duke Quantum Center, the NC State Quantum Initiative, and the UNC Kenan-Flagler's Rethinc. Labs.