Demonstration of a fast π-2π-π Rydberg entangling gate for finite blockade

We describe an entangling gate for two atomic qubits based on the interaction between Rydberg states. This protocol can produce any desired two-qubit controlled-phase gate when combined with single-qubit rotations. The gate adapts the approach of the classic π-2π-π Rydberg-blockade entangling gate to the finite-blockade regime. Rydberg excitation pulses are applied sequentially to the control, target, and control atoms, respectively. The first (third) pulse transfers population in one of the control atom’s qubit states to (back from) a Rydberg level. The second pulse drives a qubit-to-Rydberg Rabi oscillation from one of the target atom’s qubit states. The detuning of this oscillation depends on whether the control atom was transferred to the Rydberg level, which would impose a shift on the target atom’s qubit-to-Rydberg transition frequency. One can engineer the target atom’s qubit-to-Rydberg Rabi oscillations so that the target atom always returns to the qubit state through appropriate choice of the detuning, frequency, and duration of the oscillation.  

 

We demonstrate this gate protocol in a neutral atom quantum processor based on an array of Cesium atom clock-state qubits individually trapped in optical tweezers. For gate characterization, we use a square array of four qubits. We calibrate the gate parameters and additional single-qubit rotations to realize controlled-Z gates for each nearest-neighbor pair. These gates are implemented with a ratio of blockade strength to target atom qubit-Rydberg Rabi frequency of about 1.2. We perform quantum circuits to produce entangled Bell states, and we observe a maximum (mean) fidelity over the four qubit pairs of 92.8 (6) % (91.8 (3) %). Accounting for the measured fidelity of initial qubit-state preparation, we infer a maximum (mean) controlled-Z gate fidelity across qubit-pairs of 96.4 (7) % (95.3 (4) %). This protocol relaxes requirements on both blockade strength and addressing symmetry for high-fidelity entangling gates between neutral-atom qubits.