The quantum computer is a dream but it is not at all obvious that the current prototypes exceed one day the stage of mere curiosity and can compete with conventional computers. A group of researchers from Yale University has made a new element in building a real quantum processor.

The concept of quantum computer and, more generally, information and quantum computing, arose during the 1980s through the thinking of researchers like Richard Feynman and David Deutsch. It was then gradually realized that some calculations with classical computers could be done more efficiently, and above all much more quickly, with computers using the superposition of quantum states (with the interferences between amplitudes of probabilities for quantum systems).

One can indeed show that with such a computer, one can work out very effective algorithms to carry out certain tasks, exceeding those of the traditional computers then.

The most famous example is probably that of Peter Shor algorithm for factoring a large integer. Thus, although it might take many billions of billions of years the time needed to factorize a thousand digit figures with current computers, this operation will require about twenty minutes with a quantum computer.

Similarly, Lov Grover gave an algorithm bearing his name that allows a quantum computer to quickly find information in a database in any order.

Lov Grover – Bell Labs

A quantum computer, instead of using bits of information, would use qubits. While a bit is in one of two possible binary states, represented by 0 or 1, qubits is a quantum superposition of these states. This allows a kind of calculations in parallel and therefore a quantum computer is theoretically more efficient to search for information in a database.

Robert J. Schoelkopf, a physicist at Yale University and his colleagues recently published a paper in Nature that described the realization of a superconducting circuit containing billions of atoms, capable of performing calculations with two qubits of information . The researcher gave the subject the following interesting image.

If you have 4 phone numbers which we assume one is known to you, the phone numbers should be tested on average two to three times before finding the right person. A quantum computer operating with the proper algorithm would carry out these tests, not sequentially but simultaneously. More precisely, the time necessary to find information then grows like the square root of the N number of cases to be examined rather than linearly. It is understood that one saves time and especially since we take a large number of phone numbers into account.

A real circuit which provides a gain by a factor of miles.

However, to beat a classical computer, a quantum model should include a large number of qubits, which usually implies a large number of quantum systems, such as particles on the qubits in a state of superposition.

Unfortunately, when this number grows, it approaches the behavior of a traditional system. It seems like the famous phenomenon of decoherence solving the paradox of Schrödinger’s cat.

The object able to behave as a quantum computer seems to be able to do it only during a time as much shorter than its power is large. In the present state of knowledge, it is betting big that the effects of decoherence can not be sufficiently controlled to enable the realization of a quantum computer of good size, able to beat only in speed and power of  a classic supercomputer at least in the case of certain algorithmic problems.

It has so far managed to achieve tiny quantum computers operating with ions, atoms or photons. But this is the first time one produces a true quantum processor based on billions of atoms.

While a decade ago the effects of quantum superpositions could be maintained only for a nanosecond or so, researchers now manage with this device pertaining to the solid state physics to achieve a quantum coherence time of a microsecond. It is a thousand times more..

The road ahead is still long before beating a computer as Roadrunner and it is not at all sure that this is possible one day, because of the effectiveness of decoherence illustrated by the work of Serge Haroche.

Richard Feynman in a seminar at CERN in 1965 just after receiving his Nobel Prize. He is one of the fathers of the theory of quantum computers.

Quantum computer 20 years off