In this new article we used our NMR quantum computer (only two qubits…), in collaboration with Frederico Brito (USP-SC), to simulate in a digital-analog way a quantum annealing process. We currently hold (as far as I know…) the record of Trotter steps (235) and gates (more than 2000) in a quantum simulation. Moreover, we were able to relate the quality of the computation with the amount of entanglement it generated. But don’t let this trick you: more entanglement does seem to relate with better computation quality, but it does not necessarily imply better-than-classical computation! See the plot and description below, and click here for the full article.
Title: Reliability of digitized quantum annealing and the decay of entanglement
Authors: John P. S. Peterson (CBPF), Roberto S. Sarthour (CBPF), Alexandre M. Souza (CBPF), Ivan S. Oliveira (CBPF), Frederico Brito (USP-SC), Fernando de Melo (CBPF)
Abstract: We performed a banged-digital-analog simulation of a quantum annealing protocol in a two-qubit Nuclear Magnetic Resonance (NMR) quantum computer. Our experimental simulation employed up to 235 Trotter steps, with more than 2000 gates (pulses), and we obtained a protocol success above 80%. Given the exquisite control of the NMR quantum computer, we performed the simulation with different noise levels. We thus analyzed the reliability of the quantum annealing process, and related it to the level of entanglement produced during the protocol. Although the presence of entanglement is not a sufficient signature for a better-than-classical simulation, the level of entanglement achieved relates to the fidelity of the protocol.