高清福利片

It鈥檚 no surprise that quantum computing has become a media obsession. A functional and useful quantum computer would represent one of the century鈥檚 most profound technical achievements.

For researchers like me,听听is welcome, but some claims appearing in popular outlets can be baffling.

A recent听听补苍诲听听from the tech giants has woken the interest of analysts, who are now eager to proclaim a breakthrough moment in the development of this extraordinary technology.

Quantum computing is described as 鈥渏ust around the corner鈥�, simply awaiting the engineering prowess and entrepreneurial spirit of the tech sector to realise its full potential.

What鈥檚 the truth? Are we really just a few years away from having quantum computers that can听? Now that the technology giants are engaged, do we sit back and wait for them to deliver? Is it now all 鈥渏ust engineering鈥�?

Why do we care about quantum computing?

Quantum computers are machines that use the rules of听听鈥� in other words, the physics of very small things 鈥� to听听in new ways.

They exploit the unusual physics we find on these tiny scales, physics that defies our daily experience, in order to solve problems that are exceptionally challenging for 鈥渃lassical鈥� computers. Don鈥檛 just think of quantum computers as faster versions of today鈥檚 computers 鈥� think of them as computers that function in a totally new way. The two are as different as an abacus and a PC.

They can (in principle) solve hard, high-impact questions in fields such as codebreaking, search, chemistry and physics.

Chief among these is 鈥渇actoring鈥�: finding the two prime numbers, divisible only by one and themselves, which when multiplied together reach a target number. For instance, the prime factors of 15 are 3 and 5.

As simple as it looks, when the number to be factored becomes large, say 1,000 digits long, the problem is effectively impossible for a classical computer. The fact that this problem is so hard for any conventional computer is how we secure most internet communications, such as through听.

Some quantum computers are known to perform factoring exponentially faster than any classical supercomputer. But competing with a supercomputer will still require a pretty sizeable quantum computer.

Money changes everything

Quantum computing began as a unique discipline in the late 1990s when the US government, aware of the newly discovered potential of these machines for codebreaking, began听

The field drew together teams from all over the world, including Australia, where we now have听听听in quantum technology (the author is part of of the Centre of Excellence for Engineered Quantum Systems).

But the academic focus is now shifting, in part, to industry.

IBM has long had a听听in the field. It was recently joined by Google, who听, and Microsoft, which has partnered with academics globally, including听.

Seemingly smelling blood in the water, Silicon Valley venture capitalists also recently began investing in new听听working to build quantum computers.

The media has mistakenly seen the entry of commercial players as the genesis of recent technological acceleration, rather than a听response听to these advances.

So now we find a variety of competing claims about the state of the art in the field, where the field is going, and who will get to the end goal 鈥� a large-scale quantum computer 鈥� first.

The strangest of technologies

Conventional computer microprocessors can have more than one billion fundamental logic elements, known as transistors. In quantum systems, the fundamental quantum logic units are known as qubits, and for now, they mostly number in the range of a dozen.

听are exceptionally exciting to researchers and represent huge progress, but they are little more than toys from a practical perspective. They are not near what鈥檚 required for factoring or any other application 鈥� they鈥檙e too small and suffer too many errors, despite what the frantic headlines may promise.

For instance, it鈥檚 not even easy to answer the question of which system has the best qubits right now.

Consider the two dominant technologies. Teams using听听听have qubits that are听, but relatively slow. Teams using听听qubits (including听听补苍诲听) have relatively error-prone qubits that are much faster, and may be easier to replicate in the near term.

Which is better? There鈥檚 no听. A quantum computer with many qubits that suffer from lots of errors is not necessarily more useful than a very small machine with very stable qubits.

Because quantum computers can also take different forms (general purpose versus tailored to one application), we can鈥檛 even reach agreement on which system currently has the greatest set of capabilities.

Similarly, there鈥檚 now seemingly endless competition over simplified metrics such as the number of qubits.听, 16,听! The question of whether a quantum computer is useful is defined by much more than this.

Where to from here?

There鈥檚 been a media focus lately on achieving 鈥溾��. This is the point where a quantum computer outperforms its best classical counterpart, and reaching this would absolutely mark an important conceptual advance in quantum computing.

But don鈥檛 confuse 鈥渜uantum supremacy鈥� with 鈥渦tility鈥�.

Some quantum computer researchers are听听slightly arcane problems that might allow quantum supremacy to be reached with, say, 50-100 qubits 鈥� numbers reachable within the next several years.

Achieving quantum supremacy does not mean either that those machines will be useful, or that the path to large-scale machines will become clear.

Moreover, we still need to figure out how to deal with errors. Classical computers rarely suffer hardware faults 鈥� the 鈥渂lue screen of death鈥� generally comes from software bugs, rather than hardware failures. The likelihood of hardware failure is usually less than something like one in a听, or听10^-24听in scientific notation.

The best quantum computer hardware, on the other hand, typically achieves only about one in听, or 10^-4.听That鈥檚 20听orders of magnitudeworse.

Is it all just engineering?

We鈥檙e seeing a slow creep up in the number of qubits in the most advanced systems, and clever scientists are thinking about听听that might be usefully addressed with small quantum computers containing just a few hundred qubits.

But we still face many fundamental questions about how to build, operate or even validate the performance of the large-scale systems we sometimes hear are just around the corner.

As an example, if we built a fully 鈥溾�� quantum computer at the scale of the millions of qubits required for useful factoring, as far as we can tell, it would represent a totally new state of matter. That鈥檚 pretty fundamental.

At this stage, there鈥檚 no clear path to the millions of error-corrected qubits we believe are required to build a useful factoring machine.听听(in which this author is a participant) are seeking to build just one error-corrected qubit to be delivered about five years from now.

At the end of the day, none of the teams mentioned above are likely to build a useful quantum computer in 2017 鈥� or 2018. But that shouldn鈥檛 cause concern when there are so many exciting questions to answer along the way.

This article was first published in .