Schedule path optimization for quantum annealing and adiabatic quantum computing

摘要: Adiabatic quantum computing and optimization have garnered much attention recently as possible models for achieving a advantage over classical approaches to other special purpose computations. Both techniques are probabilistic in nature the minimum gap between ground state first excited of system during evolution is major factor determining success probability. In this work we investigate strategy increasing probability by introducing intermediate Hamiltonians that modify path initial final Hamiltonians. We focus on an problem relevant recent hardware implementations present numerical evidence existence purely local Hamiltonian achieve optimum performance terms pushing one end points evolution. As part study develop convex formulation search optimal adiabatic schedules makes computation more tractable, which may be independent interest. further effectiveness random probability, empirically find significant but only modest amount.