Southwest Quantum Information and Technology

Improving optimal control in a cold-atom qudit analog quantum simulator

Presenting Author: Kevin Kuper, University of Arizona
Contributing Author(s): Nathan Lysne, Poul Jessen, Pablo Poggi, Karthik Chinni, Ivan Deutsch

As noisy intermediate-scale quantum devices become commonly realized, it is unclear how useful they will be when attempting classically hard problems such as analog quantum simulation (AQS) in the presence of imperfections. We aim to understand the limitations of such devices by studying the behavior of our own small-scale, high fidelity quantum processor. Operating in the 16-dimensional hyperfine ground manifold of a single Cs atom, this system is controllable with a combination of rf and µw magnetic fields, each subject to arbitrary piece-wise linear phase modulation. Gradient-ascent optimal control allows us to find a set of phase modulation control waveforms that implement any arbitrarily-chosen unitary map with high accuracy. This property of our general-purpose simulator allows us to not only simulate any given system in Hilbert space dimensions up to 16, but also to do so in an arbitrary basis. If we allow this basis to be unconstrained when using our optimal control protocol, we reduce the number of control parameters and obtain both shorter control times and higher experimental fidelities. However, there remain classes of unitaries for which it is difficult to find high-performing control waveforms and to simulate for long times while maintaining good fidelity. One important factor could be the presence or absence of degeneracies in the spectrum of the model Hamiltonian. Further study of these limitations may illuminate how errors can impact AQS of complex dynamics.

(Session 5 : Sunday from 5:00pm - 7:00pm)