Quantum computers seem to be very much more powerful than traditional computers. For example, integers can be factored efficiently on a quantum computer but probably not on a traditional computer. However, quantum computers are not yet practical. A 7 bit quantum computer has been build by IBM, and some quantum algorithms have been run on it, but the experimental results are very far behind theoretical results.
A course lecture at Indiana University summarized the outlook for quantum computers in the near future. I will presently summarize the discussion given in that lecture.
There are many issues that have to be overcome in order to have a practical quantum computer. A quantum computer must be built that has individually addressable qubits and individually controllable couplings. In addition, the amount of time that a quantum register is usable must be increased so that a computation of the order of at least thousands or tens of thousands of gates can be accomplished. Currently, quantum registers do not last long enough for this to be accomplished. In order for a quantum register to be the basis for fault-tolerant computation, it would have to have at least on the order of 13 physical qubits per one logical qubit. Ultimately, the quantum register would have to consist of 416 physical qubits in order to have 32 logical bits (the size of current computer registers). Current technology is too limited to be able to deliver these kind of results, and thus there will be no fault-tolerant registers.
Quantum computation may be carried out in the near future on raw qubits and on increasingly large registers. However, error control would have to be based on working with statistical ensembles of registers. NMR is the technology that will be used to do this. The author believes that NMR will continue to grow but ultimately has limits.
In addition, there needs to be research in the area of automating quantum computations. One possibility is to develop compilers, which, given a circuit diagram would generate the pulses required to implement the circuit. The author also states that a molecule should be designed specifically for quantum computing that could contain the large number of qubits that are required.
As already stated, the computation time must be extended in order to yield practical computation power. However, the author states that there is a limit to the usefullness of quantum computers using current technology. He envisions research continuing with current technology in order to build practical experience with quantum computation. By the time that techniques have been developed for quantum computation, there will likely be a new implementation of qubits that will allow the leap to full practicality. A variety of possibilities are listed for this implementation will be: quantum dots, Josephson juctions, nuclear spins embedded in crystal lattice, and anyons.
There are many hurtles that must be overcome in terms of hardware. However, there are also many advances that need to be made in software. There are currently only a few quantum algorithms that could be useful. Quantum computers will not be useful until there are a wide range of useful algorithms designed for them, so research must be done in designing algorithms for quantum computers.
There are a broad range of areas that must be developed in order for quantum computers to be practical, but the potential payoff is great. The NSF has recognized this potential by providing special funding this year for certain areas related to quantum computing.
Further information:
IBM's 7 bit quantum computer.