This paper shows that a widely used physics model — the transverse-field Ising model with a single, time-dependent global field — can simulate the standard gate model of quantum computation, provided the global field is allowed to vary non-monotonically in time. In plain terms: by carefully pulsing a single, shared control field while using fixed local couplings, the authors construct a way to perform any quantum circuit with only polynomial overhead in qubit number, time and energy scale.

The transverse-field Ising model is a simple quantum model often used in analog devices and quantum annealers. It combines a classical energy term (an Ising Hamiltonian that couples pairs of qubits) with a transverse field that induces quantum flips. The authors consider the case where the transverse field is the same on every qubit and can be driven as a function Γ(t) of time. Building on a prior method by Cesa and Pichler for globally controlled atoms, they use a non-monotonic schedule for Γ(t) together with static local controls to generate effective quantum gates on groups of qubits. Strong, fixed couplings let logical information be moved and processed across the device.

Why this matters: proving such a device can simulate the gate model establishes a form of “universality.” It places many analog platforms that implement a transverse-field Ising Hamiltonian — including devices from D-Wave, Pasqal and QuEra mentioned by the authors — into the same theoretical class as gate-based quantum computers. The work also has an implication in complexity theory: if quantum computers are strictly more powerful than classical ones, then there can be no efficient classical algorithm that simulates the time-dependent dynamics of this global transverse-field Ising model. In other words, the result becomes a no-go theorem for classical simulability under that common complexity assumption.