This paper presents a technique for manipulating quantum information stored in collective states of mesoscopic ensembles of cold atoms. The method uses optical excitation into states with strong dipole-dipole interactions, which leads to a "dipole blockade" effect. This effect inhibits transitions into states with more than a single excitation, enabling the controlled generation of collective atomic spin states and non-classical photonic states. It also allows for scalable quantum logic gates. The technique is demonstrated using a cold Rydberg gas. The dipole blockade is a phenomenon where the strong dipole-dipole interactions between atoms block transitions into multi-excited states. This effect is similar to mesoscopic effects in nanoscale solid-state devices. Combined with the exceptional control and long coherence times typical of quantum optical systems, this allows for the implementation of various quantum processing protocols with reduced experimental requirements. The paper discusses the use of collective atomic states for quantum information processing, including the generation of superpositions of collective spin states, coherent conversion between collective spin states and photon wavepackets, and quantum gate operations between distant qubits. Potential applications include sub-shot noise spectroscopy, secure cryptography, and scalable quantum logic devices. The basic element of the scheme is an ensemble of N identical multi-state atoms. The atoms can be trapped and prepared in specific sub-levels of the ground state. Relevant states include storage states for long-time qubit storage and long-lived Rydberg states. The manipulation of atoms is done using light fields of different frequencies and polarizations. The interaction between collective atomic excitations is discussed, with a focus on resonant dipole-dipole interactions between Rydberg atoms. The Hamiltonian describing these interactions is written in terms of collective operators. The dipole blockade effect leads to a splitting of the excited Rydberg components, which can be used to suppress resonant excitations from singly to doubly-excited states. The paper also discusses the interaction of many-atom systems with light, focusing on the creation of single collective qubits. The evolution of the system is described using effective two-level dynamics. The system can be driven into superpositions of collective states, and single-quantum excitations can be stimulated into storage sub-levels. The dipole blockade mechanism also facilitates quantum logic operations. For example, qubits stored in few-micron spaced atomic clouds can be entangled if the transitions in one ensemble are inhibited when a collective Rydberg state is excited in a second ensemble. This scheme utilizes the same principle as that of Ref. [15], with single-particle excitations replaced by collective qubits. The paper also discusses practical issues associated with the technique, including decoherence of collective excitations and the robustness of symmetric entangled states. The technique is applicable to a variety of interacting many-body systems, ranging from trapped ions to semiconductor structures. The paper concludes that the dipole blockade can be used for coherent manipulation and entanglement ofThis paper presents a technique for manipulating quantum information stored in collective states of mesoscopic ensembles of cold atoms. The method uses optical excitation into states with strong dipole-dipole interactions, which leads to a "dipole blockade" effect. This effect inhibits transitions into states with more than a single excitation, enabling the controlled generation of collective atomic spin states and non-classical photonic states. It also allows for scalable quantum logic gates. The technique is demonstrated using a cold Rydberg gas. The dipole blockade is a phenomenon where the strong dipole-dipole interactions between atoms block transitions into multi-excited states. This effect is similar to mesoscopic effects in nanoscale solid-state devices. Combined with the exceptional control and long coherence times typical of quantum optical systems, this allows for the implementation of various quantum processing protocols with reduced experimental requirements. The paper discusses the use of collective atomic states for quantum information processing, including the generation of superpositions of collective spin states, coherent conversion between collective spin states and photon wavepackets, and quantum gate operations between distant qubits. Potential applications include sub-shot noise spectroscopy, secure cryptography, and scalable quantum logic devices. The basic element of the scheme is an ensemble of N identical multi-state atoms. The atoms can be trapped and prepared in specific sub-levels of the ground state. Relevant states include storage states for long-time qubit storage and long-lived Rydberg states. The manipulation of atoms is done using light fields of different frequencies and polarizations. The interaction between collective atomic excitations is discussed, with a focus on resonant dipole-dipole interactions between Rydberg atoms. The Hamiltonian describing these interactions is written in terms of collective operators. The dipole blockade effect leads to a splitting of the excited Rydberg components, which can be used to suppress resonant excitations from singly to doubly-excited states. The paper also discusses the interaction of many-atom systems with light, focusing on the creation of single collective qubits. The evolution of the system is described using effective two-level dynamics. The system can be driven into superpositions of collective states, and single-quantum excitations can be stimulated into storage sub-levels. The dipole blockade mechanism also facilitates quantum logic operations. For example, qubits stored in few-micron spaced atomic clouds can be entangled if the transitions in one ensemble are inhibited when a collective Rydberg state is excited in a second ensemble. This scheme utilizes the same principle as that of Ref. [15], with single-particle excitations replaced by collective qubits. The paper also discusses practical issues associated with the technique, including decoherence of collective excitations and the robustness of symmetric entangled states. The technique is applicable to a variety of interacting many-body systems, ranging from trapped ions to semiconductor structures. The paper concludes that the dipole blockade can be used for coherent manipulation and entanglement of