Date: 2008-08-07

Proc. Natl. Acad. Sci. USA 2008 105:11050-11054; published ahead of print August 5, 2008,

doi:10.1073/pnas.0800740105

doi:10.1073/pnas.0800740105

Quantum computers that make use of the principles of quantum mechanics are expected to solve problems such as factoring large integers,database search and quantum physics simulation much faster than any classical computer. However, building such devices in practice has proved extremely difficult. A significant challenge is that the unavoidable coupling of the quantum computers to the environment quicklydestroys the fragile quantum information. It is thus important to find ways to reduce the decoherence and carry out coherent quantum operationsin the presence of noise. A major source of the noise is that the physical qubits can be lost during quantum computing. The fundamental unit for quantum computingthe qubit, is supposed to be an isolated two- level system consisting of a pair of different quantum states. However, for most proposed quantum hardware, the qubits could leak out of the desired qubit space and lost into a larger Hilbert space. Take, single photons, for example; they canbe lost during processing or owing to inefficient photon sources anddetectors. This causes loss

of quantum information, posing a significantobstacle for practical quantum computation.

Fortunately, it is possible to tackle the problem by designing appropriate codes, which are usually called as “loss-tolerant quantum code”. A group of physicists led by Dr. Jian-Wei Pan from the Universitof Science and Technology of China and the University of Heidelberg has experimentally demonstrated such codes for the first time. It is shown that by grouping together four photons to encode the logical quantuminformation of a single photon, even if any one of the four photons is lost, the essential quantum information is still preserved and can brecovered from the remaining three photons. This result proves that the qubit loss error could in principle be overcome by quantum codes, thus representing a necessary step toward scalable quantum informatioprocessing.

reference: Lu et al. Proc. Natl. Acad. Sci. 105, 11050-11054 (2008).