Reporter Adam Levy talks to Andrew King about D-Wave’s latest research1 with quantum simulators.

Subscribe to the Nature Podcast on iTunes or your favourite podcast app.

Head here for the Nature Podcast RSS feed.

TRANSCRIPT

Interviewer: Adam Levy

Picture a computer. Now I’m going to go out on a limb and say that you were probably thinking of your laptop, or maybe even your smartphone if you’re being smart about it. There’s a good chance you weren’t thinking of a quantum computer. These hulking, supercooled machines can take up the best part of a room, but still aren’t capable of doing anything that a regular computer can’t. Even so, there’s something fundamental about quantum computers that could one day give them an advantage…

Interviewee: Andrew King

A bit which is the fundamental building block of any computer program is a binary digit, which is just a 0 or a 1.

Interviewer: Adam Levy

This is computer scientist, Andrew King. But the building blocks of quantum computers aren’t bits. Instead quantum computers use qubits. Qubits can be composed of the spin of an electron or a current in a superconducting loop. But whatever is used to make a qubit, the crucial thing is how they behave:

Interviewee: Andrew King

A qubit is a 0 or 1 or both at same time. The probabilities of 0 and 1 occurring interact in ways that are not possible in classical computing.

Interviewer: Adam Levy

And this otherwise impossible interaction between qubits means quantum computers can perform otherwise impossible computations. At least that’s the theory. Here’s Davide Venturelli — also a computer scientist.

Interviewee: Davide Venturelli

Well we cannot certainly claim — at least to my knowledge – that something that has not been done with a classical computer has been done with a quantum computer. Yet.

Interviewer: Adam Levy

Now, spoiler alert, we’re not bringing you news of this so-called ‘quantum supremacy’ — a quantum computer performing a task no regular computer can handle. But a paper out this week does offer an important step.

You see, one of the reasons that quantum computers are yet to achieve quantum supremacy and leapfrog classical computers is that they are just really hard to build. Most quantum computers only have few handfuls of qubits — not enough to handle tasks that your laptop can’t. But one company, D-Wave, has built a machine with some 2000 qubits.

Interviewee: Davide Venturelli

In my opinion, D-Wave really schooled the world that if you are really determined to create a quantum computer, to build a quantum computer, you can. What they did was create a processor that exploits some effects of quantum mechanics. Not all of the effects and all of the power we would want, but let’s say a minimum viable quantum computer. But then the challenge will be to be able to create algorithms and procedures and computational methods that can work with the constraints of this architecture. So it’s a double edged sword.

Interviewer: Adam Levy

So these D-Wave machines may have more qubits than other quantum computers, but they’re also trickier to work with. Andrew, who we heard from earlier, works at D-Wave. He set out to simulate something that quantum computers should naturally have an advantage at:

Interviewee: Andrew King

So back in the early 80s, Richard Feynman proposed the idea of a quantum simulator where you would want to study a quantum system. And even though you know the equations that describe this system, they’re very difficult to solve. But if you can somehow program another quantum system to simulate the quantum system that you’re interested in, then you can get a big advantage over the classical simulation that you’d have to do.

Interviewer: Adam Levy

In other words, because a quantum computer is already, well, quantum, it can simulate certain quantum systems more naturally than a normal computer.

Interviewee: Andrew King

What we were trying to investigate with this experiment is the onset of what’s called a topological phase transition. So it’s essentially a very exotic phase of matter in a magnetic lattice. And so physicists have studied this system before and we wanted to see if we could make our computer instantiate this system.

Interviewer: Adam Levy

To do this, the team had to set up the computer so that it would simulate the system correctly. But they also had to find a way to read out the answer. And this isn’t a trivial task like reading something off your laptop screen. The team have to peer at what the delicate qubits are doing in the middle of the process.

Interviewee: Andrew King

It’s all really difficult because the system is extremely sensitive, and so balancing it was a big challenge. You can imagine trying to present somebody a marble on a plate. It’s really difficult to do unless you have the plate balanced very well.

Interviewer: Adam Levy

But Andrew was able to balance the plate. The D-Wave computer was able to simulate the topological phase transition in a material, getting the same results as a classical computer. Of course, the fact the team could simulate this problem using a classical computer at all indicates this is not an example of ‘quantum supremacy’. But while we might not be at a computing sea-change just yet, Andrew says the quantum computer simulation did still outperform its classical counterpart.

Interviewee: Andrew King

Yeah, the results were very good. Our results indicated that we are much, much faster. So at least thousands of times faster. The ultimate hope is that we can study materials efficiently without making them in a lab.

Interviewer: Adam Levy

With this work, a practical use for quantum computers might be that much closer. These machines which have been worked on for decades may finally be moving from theoretical playthings to actual useful research tools, simulating materials and processes that would be impossible to study on conventional hardware. For Davide, who didn’t work on the study, this is a big deal.

Interviewee: Davide Venturelli

I think the paper by Andrew King and collaborators is a very, very, very interesting development because the level of analysis is unprecedented. And I think that it will pave the way towards more experiments to be able to represent physical models, which are highly non-trivial.

Interviewer: Adam Levy

But pushing the envelope is only part of the motivation for Andrew. And he looks forward to more and more clever computations with these quantum machines.

Interviewee: Andrew King

I have to admit that half the pleasure of it is just working on a really cool problem and having things work out the way that you wanted them to. But I think that we’re kind of at a turning point with quantum computing. The competitive playing field is getting a lot more crowded and I think we’re going to show some really interesting results — and other people will show some really results — in the next few years.