Two Steps Forward for Quantum Computers

Two sets of researchers into quantum computing have announced very different sets of results. One claim they've already achieved more powerful performance than a traditional computer, while the others say they've confirmed the potential for performing genuinely useful tasks.

Quantum computing aims to take advantage of the way particles can exist in more than one state at a time. That compares with traditional computers which store data in bits, each of which is either a 0 or a 1. In principle at least, a quantum computer could be made up of "qbits" and perform calculations dramatically faster by using the same qbits to represent multiple data points at once.

Quantum Supremacy

The way this works means adding only a few extra qbits dramatically improves performance. Google notes that in the past five years it's worked on a quantum computer that's increased from 53 qubits to 70 qubits. That increase resulted in a machine 241 million times more powerful. (Source: fudzilla.com)

Google now claims to have achieved "quantum supremacy," meaning it has produced a working quantum computer that outperforms even the faster traditional computer on a particular task.

It's a symbolic moment, but critics say the tasks used for such speed test have few real world uses. However, IBM and a University of Zurich researcher say they've now produced a quantum computer that increases the speed of a "Monte Carlo simulation."

Speedier Simulations

A Monte Carlo simulation is a computer operation that models the effects of random variables and then looks at the probability of different outcomes.

A simple example would be modeling the NBA playoffs a thousand times and seeing which team won out most often. A more complex example is modeling weather for the next 24 hours, which is why weather presenters will usually forecast the likelihood of rain as a percentage, rather than make an outright prediction of whether or not it will rain.

Such calculations usually requires extremely powerful supercomputers, particularly as they cover more variables. That's because it's a slow process going through every simulation one step at a time. Being able to effectively run multiple simulations at once dramatically cuts the total time.

The researcher say the exciting discovery is that while more complex and detailed simulations still increase the total processing time with quantum computers, the rate of increase is slower than with traditional computers. (Source: newscientist.com)

We're still a long way from the point where the quantum computer is the better option for a Monte Carlo simulation. The researchers believe the cross-over point would be around 1,000 qubits. However, from that point the quantum computer could do a much better job at many real-world tasks such as mapping the effects of natural disasters, mapping public transport routes more efficiently or finding the least wasteful way to arrange industrial processes.

What's Your Opinion?

Can you get your head around quantum computing? Is it a worthwhile research area? Are practical benefits more important than simply getting faster speeds?