Bringing quantum physics into computing would make it possible to rapidly develop algorithms which today would take many years to build. Google has now entered the fray, hiring for its ongoing project one of the foremost brains in quantum computing.
In order to speed up research at Quantum Artificial Intelligence Labs – an initiative launched by NASA and Google in partnership with a consortium of universities – Google has now brought on board Physics Professor John Martinis and engaged the services of his entire quantum computing team at the University of California, Santa Barbara. This move marks an opportune moment for us to take a look at how research in this field is progressing, as some in the scientific community are hoping that Google’s involvement in building a quantum computer will send encouraging signals to the computing industry. Back in 2012, Amazon CEO Jeff Bezos and the CIA’s investment arm In-Q-Tel illustrated the growing wave of interest in building a new generation computer when they pumped investment dollars into Canadian startup D-Wave Systems. Google’s stated aim is to take a steady approach to developing quantum computers, but the Mountain View giant also plans to work with D-Wave's scientists as D-Wave scales up to a 1,000-qubit* (quantum bit) processor.
New paradigm set to speed up computing power?
For the last eight years D-Wave Systems has been financed by a variety of sources, from Jeff Bezos to the CIA-run In-Q-Tel, which were part of a $30 million funding round. Since 2011 Google has been tinkering with a D-Wave computer but nevertheless showed some caution before finally launching wholeheartedly into the quantum computing adventure. This now looks to be a race against time between the US Internet giants to take computing power to the next level. For several years now the potential of quantum computing has been creating a stir in the scientific community and the 2012 Nobel Prize in Physics was awarded to two scientists working in the field. Meanwhile Berkeley, California-based Rigetti Computing, a graduate of the Y-Combinator startup incubator, is working to commercialise quantum processors over the next few years. At the same time, Washington DC-based quantum computing software and services startup QxBranch is currently participating in Accenture's FinTech Innovation Lab in Hong Kong – a programme designed to nurture early-stage start-ups from around the world that are developing new technologies for the financial services sector, with a particular focus on the Asia-Pacific region. However, while the still-strictly-theoretical promise of quantum computing may lead to the development of artificial intelligence at levels never seen before, such very basic questions as the architecture of a quantum computer still remain to be answered.
Google drawing together rival scientific approaches to quantum computing
Despite all the efforts made so far, the intrinsic lack of stability in quantum manipulation makes it difficult to build a central processing unit which can achieve reliable fast computing. Whereas digital computers require data to be encoded into binary digits (bits), each of which is always in one of two definite binary states – i.e. 0 or 1, off or on – quantum computation uses quantum bits (qubits*), which can basically exist in both states – or rather be ‘suspended’ between the two states – for a fleeting few microseconds, enabling two or more calculations to be carried out simultaneously. D-Wave Two, which was acquired for the Quantum Artificial Intelligence Lab, is a 512-qubit ‘quantum annealing’ computer designed for research into machine learning, among other fields of study, which can however only solve optimisation problems rooted in a number of variable factors, such as working out the best itinerary to take. These limitations have now prompted Google to hire the services of John Martinis and his team. Before joining Google, the team was developing error correction techniques that can stabilise quantum bits. Professor Martinis’ laboratory in Santa Barbara had previously built quantum computing systems of up to nine qubits based on superconducting quantum circuits, with a non-significant error rate. Being able to manipulate five or more qubits means the quantum computer is as good as today’s science can make it. However, in order to use the true potential of quantum computing for the purpose of genuine scientific discoveries, a central processing unit would need to be able to manipulate many thousands of qubits.