Quantum computing: far from practical

Quantum computing is a fascinating and exciting field, but one with limited real-world impact at present. We are still a long way from a working quantum computer. Even when we do have working models, you aren't likely to want one: quantum computing is more than just regular computing done faster.

IBM's breakthroughs certainly are noteworthy, though, and deserve all the plaudits they have received. But they don't really have any implications for the world of everyday computing.

The basic building block of quantum computing is the qubit. Unlike computing's traditional bits, which hold specific and unique values of 0 or 1, qubits can hold both values simultaneously, in quantum superposition. Multiple qubits, therefore, can represent a massive number of simultaneous values.

Although the state-of-the-art has been sufficient to create these building blocks of quantum computing for a few years, these blocks have been sufficient only to demonstrate their potential, like a prototype car, which achieves astounding speed, but then immediately breaks down.

Qubits hold multiple potential values, but only until their quantum state collapses, a process known as decoherence, at which point a single valid state remains. Any interaction with the external world (“interference”) can cause the quantum state of the qubit to collapse, and the more qubits involved, the more error-prone the system becomes. Once fully collapsed, the qubits represent a single value, so offer no benefit over traditional bits.

A key challenge in bringing quantum computing to practical use is that decoherence is a very short-term phenomenon, at the nanosecond scale.

IBM's first key breakthrough is a technique which extends the life of a qubit by up to four times over previous techniques, to as long as 100 microseconds. That is within touching distance of the estimated requirements for qubits to be put to practical use. IBM's physicists isolated qubits within three-dimensional cavities, where they exhibited a much greater lifespan. By comparison, in 1999, the lifespan of a qubit was about 10 000 times shorter - useful only to demonstrate that the theory was sound.

Longer-living qubits means more qubits may be coupled together into logic gates and circuits - the next steps towards a working quantum computer.

Another IBM breakthrough was the creation of a working controlled-NOT gate, built from qubits working together. A C-NOT gate is one of many quantum gates (similar to the AND/OR/NOT gates of Boolean logic), which facilitate quantum computing. A C-NOT gate flips the state of one qubit if another holds the value one - simple-sounding logic, but a core capability enabling more complicated systems.

IBM's C-NOT gate is a key practical demonstration of a fundamental component of quantum computing, taking principles out of the realm of theory and into the real world. Assembling multiple logic gates will create working logic circuits for processing quantum algorithms.

With the extended qubit lifespan, and working gates, IBM believes it is poised to assemble a working prototype of a quantum computer.

Impressive though IBM's accomplishments are, you are not going to have a quantum computer under your desk any time soon. For a start, we are many years away from a working prototype, never mind commercial availability.

Even if you could buy one, what might you do with it? Most conventional algorithms in modern computing are simply not suited to quantum computing, and would not benefit from a quantum computer versus a traditional computer.

Some do, though. There are classes of quantum algorithms which stand to benefit certain types of problems massively, such as those which are particularly suited to parallel calculation, including prime factorisation (useful for cryptography and cryptanalysis) and simulation.

It is likely that the initial beneficiaries of quantum computers will be the same ones that adopted early mainframes: governments and research institutions, and not just because of the costs and logistics. This is bit of a chicken-and-egg problem: we don't have a lot of useful real-world quantum algorithms, because we don't have a working quantum computer yet. The logic is well-researched, though.

Size is also a factor. Quantum computing benefits from superconduction, which requires very low temperatures and the concomitant bulky equipment. Of course, early IBM mainframes were room-sized beasts too, and now we pack more punch into a smartphone than they could fit in a building. Lab prototypes are more concerned with function than size, but for now there is no expected breakthrough (like room-temperature superconductors), which could shrink quantum computers to a more desktop-friendly footprint.

Lastly, don't confuse quantum computing with quantum key distribution (sometimes mistakenly called quantum encryption). Quantum key distribution uses quantum entanglement to ensure that encryption keys can be exchanged without interception. Several commercial implementations of quantum key distribution are already in the market.

Additional reading: IBM's announcement: http://www-03.ibm.com/press/us/en/pressrelease/36901.wss, Wikipedia's starting page for quantum computing concepts: http://en.wikipedia.org/wiki/Quantum_computer.

Related story:
Quantum computing to become a reality