Multiqubit Toffoli Gates and Optimal Geometry with Rydberg Atoms

Due to its potential for implementing a scalable quantum computer, the multiqubit Toffoli gate lies in the heart of quantum information processing. In this paper, we demonstrate a multiqubit blockade gate with atoms arranged in a three-dimensional spheroidal array. The gate performance is greatly improved by the method of optimizing control-qubit distributions on the spherical surface via evolutionary algorithm, which leads to an enhanced asymmetric Rydberg blockade. This spheroidal configuration, not only preserves the dipole blockade energy well between arbitrary control-target pairs, which keeps the asymmetric blockade error at a very low level, but also manifests an unprecedented robustness to the spatial position variations, leading to a negligible position error. Taking account of intrinsic errors and using typical experimental parameters, we numerically show that a c6not Rydberg gate can be created with a fidelity of 0.992, which is only limited by the Rydberg state decays. Our protocol opens up a platform of higher-dimensional atomic arrays for achieving multiqubit neutral-atom quantum computation.