Holographic Quantum Processing

点击翻译按钮获取中文摘要

This project aims at developing a new paradigm for the experimental realisation of a quantum processor. Realising a scalable quantum processor has been a long-standing goal of current international research. Experimental and theoretical research efforts have seen impressive success over recent years and superb control over small numbers of quantum bits has been demonstrated. The probably most advanced approach has been implemented with trapped ions, and calculations with a full quantum byte are possible. Quantum computers are, however, still far from everyday use and it remains a major challenge to scale such devices to large numbers of qubits and thus to technological relevance. Up to now, large amounts of high-end research technology are required for every added bit.This project will explore a potential way of circumventing this scaling of resources. Quantum bits are "traditionally" represented by individual constituents of matter, such as ions, atoms, or photons. Logical operations between qubits, required for a universal programming language, have been implemented by more or less direct interactions between these constituents. We take a different approach and represent quantum bits holographically. By taking a large amount of atoms, in this case a gas of laser cooled atoms, it is possible to encode quantum information in so-called spin waves, which are collective excitations of the gas. Here, the information is no longer stored and processed locally, but in the Fourier representation or momentum space of an ensemble. We want to experimentally demonstrate that it is possible to perform logical operations with these waves. We will assess the performance of a complete set of operations required for universal quantum computing, and thus investigate the possibility to run an arbitrary quantum algorithm on a very large number of quantum bits.We will use ensembles of ultracold atoms and employ established quantum memory techniques to prepare two classes of qubits by generating single excitations of spin waves. As in the recently demonstrated gradient echo quantum memory, excitations will be shifted in Fourier space, effectively generating two linear qubit registers like a two-taped Turing machine. Four elements are then combined to achieve universal computing capability: A) Optical scattering processes combined with single photon detectors allow for the population of the virtual Turing tapes with qubits. B) State-dependent phase-imprinting introduces the flexibility to achieve independent control over the virtual tape positions. C) Phase matching conditions enable selective optical read-out of fixed register positions. D) Microwave driving couples the virtual tapes and realises beam splitter type operations between registers. The combination of these elements leads to an unconventional implementation of the celebrated proposal by Knill, Laflamme, and Milburn for universal quantum computing with linear optics. But here, no increasing material resources are required with greater number of qubits.Ultimately, we envision a highly integrated device that links directly with fibre optics for access and communication. It will thus be compatible with applications in quantum communication which might be among the first uses of such a technology.