Quantum Computing and all that

Quantum computing has been "The next big thing" in computer design research since the mid 1980s. Companies like IBM have sunk large amounts of money into the area to develop working demonstrator prototypes of the sub-components required to build a quantum computer. Typically this involves very expensive equipment which uses super cooling technology to support qubit based calculations. Where qubits or quantum bits are based on quantum mechanical states of the polarisation of particles such as photons. The quantum mechanics aspect means that a particle can occupy two states simultaneously (otherwise known as superposition) and introduces a powerful approach to probabilistic computing.

So why all the fuss? Well there are several applications and implications. For example of a bad thing or serious implication, quantum computing computers could easily crack the cryptographics used on most public key cryptographic systems in use today. Pundits believe that it is only a matter of years before a powerful enough quantum computer is available, and then most current encryption technology becomes obsolete, as current encryption relies on the fact that it would take an impracticably long time to crack it using brute force. So there are some start up companies already offering "quantum key security" solutions to security conscious organisations.

More positively, there are other problems which become easier to address. A lot of these are modelling based. Why is this important. Well there's an old saying that "all models are wrong, but some are useful", which really means that computing based models are simplified approximations, based on assumptions and cannot always deal with real life complexity and the range of statistical variation which can occur. An interesting area is climate modelling. Almost all the climate models used today in the "climate change" debate surrounding global warming are wrong. They can't deal with all the variables and elements needed to model climate accurately and most cannot reproduce actual historical climate patterns, if they are fed historically accurate starting point data. Quantum computing could change that (provided that the input data is correct). Today's debates are fuelled by the situation where as I read once "if you put extreme data into a simple model you get extreme outputs which have nothing to do with real life". Getting a tool which accurately models climate would be extremely useful in determining what are the best actions to take to respond to long term climate change trends.

The situation is similar with economic modelling and longer term it is conceivable that quantum computing could accurately model consumer responses to marketing campaigns or even election campaigns. In engineering terms, certain numerical modelling techniques concerned with fluid dynamics, fracture mechanics and stress analysis could also be further enhanced, and the list goes onwards.

Will this change the world, probably, but we do need to think through some of the societal impacts too. Why is it interesting now? because researchers at the university of New South Wales have recently demonstrated a chip which contains all the building blocks for controlling quantum computing, using current technology. So watch this space.