Session D05: Deep Learning and Quantum Computation in Nuclear Physics

Coupled-clusters and quantum computing

This talk presents predictions from coupled-cluster calculations of rare isotopes and first results from quantum computing an atomic nucleus. \\ Rare doubly-magic nuclei play an important role because they determine the structure of entire regions in the nuclear chart. In recent years, the computation of rare isotopes such as $^{48,52,54}$Ca, $^{78}$Ni, and $^{100}$Sn and their neighbors -- based on interactions from effective field theories of quantum chromodynamics\footnote{E. Epelbaum, H.-W. Hammer, and U.-G. Meiß{\s}ner, Rev. Mod. Phys. \textbf{81}, 1773 (2009); R. Machleidt and D. R. Entem, Phys. Rep. \textbf{503}, 75 (2011); A. Ekstr{\"o}m et al., Phys. Rev. C \textbf{91}, 051301 (2015).} and using controlled approximations only\footnote{B. R. Barrett, P. Navratil, and J. P. Vary, Prog. Part. Nucl. Pays. \textbf{69}, 131 (2013); G. Hagen et al., Rep. Prog. Phys. \textbf{77}, 096302 (2014); T. A. L{\"a}hde et al., Phys. Lett. B \textbf{732}, 110 (2014); H. Herbert et al., Phys. Rep. \textbf{621}, 165 (2016).} -- led to predictions for neutron skins\footnote{G. Hagen et al., Nature Physics \textbf{12}, 186 (2016).} and the evolution of shell structure in isotopes of calcium\footnote{G. Hagen et al., Phys. Rev. Lett. \textbf{109}, 032502 (2012).}, nickel\footnote{G. Hagen, G. R. Jansen, and T. Papenbrock, Phys. Rev. Lett. \textbf{117}, 172501 (2016).}, and tin\footnote{T. D. Morris et al., arXiv:1709.02786 (2017).}. \\ Quantum computers promise to reduce the computational complexity of simulating quantum many-body systems from exponential to polynomial. Very recently, quantum computing devices have started to solve small scale, but real-world many-body problems in chemistry and magnetism\footnote{P. J. J. O'Malley et al., Phys. Rev. X \textbf{6}, 31007 (2016); A. Kandala et al., Nature \textbf{549}, 242 (2017).}. This talk presents the quantum computation of the deuteron via cloud servers\footnote{E. F. Dumitrescu et al., arXiv:1801.03897 (2018).}. This is a first step towards scalable nuclear structure computation on a quantum processor unit via the cloud, and our results shed light on how to map scientific computing applications onto nascent quantum devices.   

QCD: From the lattice to the quantum computer

A century of coherent experimental and theoretical investigations uncovered the laws of nature that underly nuclear physics — Quantum Chromodynamics (QCD) and the electroweak interactions. While analytic techniques of quantum field theory have played a key role in understanding the dynamics of matter in high energy processes, they become inapplicable to low-energy nuclear structure and reactions, and dense systems. Expected increases in computational resources into the exascale era will enable Lattice QCD calculations to determine a range of important strong interaction processes directly from QCD. However, important finite density systems, non equilibrium systems, and inelastic processes, that typically experience exponential growth in required computational resources, are expected to remain a challenge for conventional computation. There is now excitement in our community that the emergence of quantum computing may provide significant benefit in understanding these systems. In this presentation, I will discuss the state-of-the-art Lattice QCD calculations, progress that is expected in the near future, and the potential of quantum computing to address Grand Challenge problems in nuclear physics.   

Machine learning in particle physics

The use machine learning methods has become ubiquitous in particle physics experiments. I will review how these algorithms are employed, e.g. in trigger systems and data analysis, and discuss where deep learning is starting to impact our field. I will conclude by briefly looking to the future.