For nearly a quarter of a century, cutting-edge physicists have been working on building quantum computers that promise to bring a brain-warping leap in power.
Such systems would be able to tackle the most complex number-crunching calculations in seconds, rendering such things as modern encryption algorithms redundant. But now a group of scientists have shed doubt on the idea of quantum computing: what if, they argue, it's simply not possible to meet the subatomic conditions needed for quantum computers to work.
Ross Anderson and Robert Bradley of the University of Cambridge Computer Laboratory (UCCL) published their paper on the pre-print ArXiv repository. They added some additional comment on the UCCL blog.
“We’ve been assured for 25 years that quantum crypto is secure, and for 19 years that quantum computing is set to make public-key cryptography obsolete. Yet despite immense research funding, attempts to build a quantum computer that scales beyond a few qubits have failed,” wrote Anderson.
Qubits are, of course, the quantum equivalent of classical computers' bits – the mechanism for encapsulating data. The classical bit has two states – zero and one; but a qubit could be in one state or the other and indeed both at the same time. And unlike classical bits, qubits can exhibit quantum entanglement, where separated qubits are linked, such that a measurement of one will produce the same measurement on its entangled partner.
Anderson and Bradley suggest that it's possible to use classical physics to understand the entanglement process – using an experiment that bounced a small liquid drop on the surface of an oscillating liquid.
“In this two-dimensional analogue there is a limit to the number of qubits in a coherent system. It's easy to get phase coherence with waves associated with one other particle and possible to get coherence with two – one coherence per dimension,” they wrote.
They demonstrated that such a model has a good fit with the so-called Broglie-Bohm interpretation of quantum mechanics – also known as the pilot wave theory, which has been used to describe various quantum properties.
What these models suggest, conclude Anderson and Bradley, is that it will prove difficult to maintain the phase coherence of more than three qubits in a plane, or four qubits in a three-dimensional structure – unless of course we find more dimensions to work with.
But quantum computers that rely on three qubits are going to be nowhere near powerful enough to crack the types of calculations quantum computer proponents dream of.
01 Feb 2013