Thursday, January 9, 2014

Predicting quantum computers

Brian Hayes describes quantum computers and writes:
The hardware doesn’t yet exist, but languages for quantum coding are ready to go.

The year is 2024, and I have just brought home my first quantum computer. When I plug it in and switch it on, the machine comes to life with a soft, breathy whisper from the miniature cryogenic unit. A status screen tells me I have at my disposal 1,024 qubits, or quantum bits, offering far more potential for high-speed calculation than all the gigabits and terabytes of a conventional computer. So I sit down to write my first quantum program. ...

Languages such as QCL and Quipper may well solve the problem of how to write programs for quantum computers. ...

Of course all these questions remain academic until reliable, full-scale quantum computers become available. A company called D-Wave offers machines with up to 512 superconducting qubits, but the architecture of that device is not suited to running the kinds of programs generated by QCL and Quipper; indeed, there’s controversy over whether the D-Wave machine should be called a quantum computer at all. For technologies that can implement quantum circuits with controlled interference and entanglement, the state of the art is roughly a dozen qubits. With those resources, Shor’s algorithm can factor the number 21. Much work remains to be done before my kiloqubit laptop arrives in 2024.
If that is a prediction, then I would like to bet against it. Those programming languages will never run faster on a quantum computer than on the simulator. Hayes never mentions that quantum computers may be impossible, and those dozen qubits are not real.

No comments:

Post a Comment