The principal arguments for using cold, confined atoms to implement quantum computation are essentially the same as those for using cold, confined ions: long coherence times, interactions that are theoretically well-defined, high isolation of the qubits and scalability. It is reasonable to suppose that perturbations of an atomic system should be even less than in an ionic system, since ion charges interact strongly with the surrounding neutral environment, while neutral systems interact very weakly with that environment. Of course, this weak interaction between neutrals also presents challenges to the entanglement of two qubits, which is required to implement a quantum gate. Nevertheless, there has been progress both theoretically and experimentally in logic implementation.