Computer researchers from Google, working with physicists and chemists from the University of California at Santa Barbara, the University of the Basque Country, and the Basque Foundation for Science in Spain have combined two approaches to quantum computing to make the technology more useful and more suitable for commercialization than was previously possible.
Quantum computing utilizes the peculiar characteristics of quantum mechanics. In a digital computer, a bit can be either a 0 or 1, off or on. In a quantum computer, qubits can hold a 0 and a 1 simultaneously, and they can also be superimposed, creating a greater range of states and making computation faster. By taking advantage of quantum mechanical properties such as superposition and entanglement, qubits can perform certain kinds of calculations at remarkable speed, just as a GPU can outperform a CPU for specific tasks.
Such numbers-crunching agility has potential to crack encryption keys, to accelerate machine learning, and to perform complex simulations that would bring conventional systems to their knees. But it has proven to be difficult to implement.
[Can a company called D-Wave deliver on the promise of quantum computing? Read Google, NASA Bet on Quantum Computing.]
In May, IBM made quantum computing available as a service in a limited form through IBM Cloud. IBM believes that within a decade it can create a universal quantum computer with a 50- to 100-qubit processor that will surpass the most powerful conventional supercomputer available today. Its current quantum computer (digital quantum computing) is powered by a 5-qubit processor.
A Canadian company called D-Wave has a different kind of quantum computer, an analog device. It features 1,000 qubits, for about $15 million, but it is designed specifically for adiabatic quantum computing (AQC), making it useful only for specific types of calculations.
Google happens to own a D-Wave machine. Now its researchers have demonstrated an approach to quantum computing that combines the advantages of digital and analog systems.
Digital quantum computing allows for robust error correction, but it requires custom algorithms. Analog quantum computing can be generalized to deal with different algorithms more easily, but is limited by errors that accumulate in the system and by other factors.
Such errors prevent quantum computing from being applied to many of the problems computer scientists want to solve, said Google quantum engineers Rami Barends and Alireza Shabani in a blog post:
The crucial advantage for the future is that this digital implementation is fully compatible with known quantum error correction techniques, and can therefore be protected from the effects of noise. Otherwise, the noise will set a hard limit, as even the slightest amount can derail the state from following the fragile path to the solution.
In their paper, the researchers said they hope their work accelerates the development of quantum computing and encourages further research.