For the past decade, a global race has been running to build a mythical machine, the holy grail of calculations ... a quantum computer. Google recently announced they'd bought one—the D-Wave 2. Does this mean the race is over? Well ... not quite. The D-Wave 2, though costing Google around a cool $10 million, is currently only good for a certain class of problem. In Australia, teams at the University of New South Wales are working on a version that will have a much broader application ... a so-called universal quantum computer. Graham Phillips dons full clean-room regalia to check out this cutting-edge research.
NARRATOR: These guys are part of a global race. It’s been running for over a decade … to build a mythical machine, the holy grail of calculations—a quantum computer. Remarkably, first-generation quantum computers have started to appear. Indeed, earlier this year, Google bought one. The D-Wave 2.
DR ERIC LADIZINSKY: The promise of quantum computers is what would otherwise take you a billion years, you could do in a few seconds, and that’s game changing.
NARRATOR: But as impressive as the D-Wave 2 is, it can only solve certain kinds of problems.
DR GRAHAM PHILLIPS: The quantum computer these guys are working on potentially has very broad application. So the race is far from over.
NARRATOR: Conventional computers are built from silicon chips that have more than a billion miniature transistors etched on them. By cramming on ever more transistors, computers have been getting faster and faster.
PROFESSOR MICHELLE SIMMONS: So, a quantum computer looks at all possible solutions at the same time, and it gives you the right answer. So, it works in parallel. It’s that parallelism that you just don’t get with a classical computer, which really has to go one after the other. So you're expecting to get a much greater increase in computational power.
NARRATOR: In classical computers, information is stored as strings of zeros and ones. They’re called ‘bits’ and they’re represented on the chip as tiny switches that are either off or on. But the switch equivalence in a quantum computer can be in two different states at once—on and off at the same time. They’re called ‘qubits’. Research teams here, at the University of New South Wales, are working on a quantum computer based around a silicon chip with a single phosphorus atom embedded in it. A single electron from that atom serves as the qubit.
DR GRAHAM PHILLIPS: A very impressive-looking machine.
PROFESSOR MICHELLE SIMMONS: Yes. It’s basically a scanning tunnelling microscope. It’s a piece of stainless steel with all the air sucked out from the inside, so it’s an ultra-high vacuum.
NARRATOR: This machine is used to position the phosphorus atom in the silicon chip. But first, a single layer of hydrogen atoms must be added to the surface of the silicon. Then the microscope tip comes down.
DR GRAHAM PHILLIPS: So you’re going to physically knock off individual atoms with that tip.
PROFESSOR MICHELLE SIMMONS: Yeah, that’s correct. That tip, we use it to image the atoms, see where they are, and then we’ll apply a pulse above each hydrogen atom and knock it off. And literally open up a hole of exactly six atoms to let that phosphorus in.
NARRATOR: This is a world first.
PROFESSOR MICHELLE SIMMONS: We’re the only group in the world that can do it, so it’s really, you know, atomic precision to get it in there. And we find that it never behaves in the way we expect. And so you have that beautiful sense of trying to understand, right at the atomic level, what’s really happening.
NARRATOR: In this very sci-fi looking lab, another University of New South Wales team is fabricating other components needed by the quantum chip.
DR GRAHAM PHILLIPS: Knights of the Round Table.
NARRATOR: It’s a clean room.
DR GRAHAM PHILLIPS: No dust particles allowed
NARRATOR: When you’re working down at the atomic scale, a dust particle is like a boulder, so you have to be completely clean when you come in here.
DR GRAHAM PHILLIPS: Also UV light interferes with the chip-making process. So the lights here have that yellowy colour because all the UV has been filtered out.
NARRATOR: Around 100 labour-intensive steps are performed in here to build the chip. One of them is patterning a single electron transistor.
PROFESSOR ANDREW DZURAK: This is the chip. So that’s actually had the tiny patterns where the metal will go to make the single electron transistor. This has used the electron beam with just a two-nanometre spot size to write these tiny features.
DR GRAHAM PHILLIPS: Gee, that’s tiny!
PROFESSOR ANDREW DZURAK: It’s really tiny. And that transistor is the transistor that will read out the state of the spin on that atom.
DR GRAHAM PHILLIPS: This is the culmination of all that work in the clean room, a wonderful little chip there mounted on a circuit board.
NARRATOR: It contains the qubit and transistor. The researchers have chosen a phosphorus atom to make their qubit because phosphorus has one extra electron compared to the surrounding silicon.
ASSOCIATE PROFESSOR ANDREA MORELLO: Now, that one electron that’s attached to the phosphorus, like every electron, has what people call the spin. It doesn’t mean it spins on itself, it’s just an intrinsic quantum mechanical property. It’s essentially a magnetic dipole. It’s like the tiny needle of a compass.
NARRATOR: To measure the spin, the chip containing the qubit is placed inside this superconducting magnet.
ASSOCIATE PROFESSOR ANDREA MORELLO: This large magnetic field gives a different energy to the two possible spin orientations.
NARRTOR: And, in another world first, they can now detect and control the state of the phosphorus qubit in the silicon chip.
ASSOCIATE PROFESSOR ANDREA MORELLO: What you’re looking at is in real time at, you know, a thousand times per second rate at the quantum measurement of a single electron spin. In real time, before your eyes.
DR GRAHAM PHILLIPS: That is amazing! So, I mean, an electron, a tiny … infinitesimally small thing.
ASSOCIATE PROFESSOR ANDREA MORELLO: Yeah, yeah.
DR GRAHAM PHILLIPS: And we’re measuring the spin on that.
ASSOCIATE PROFESSOR ANDREA MORELLO: You’re just watching it on the computer screen. When the spike happens, it’s because the electron has left the phosphorus atom, and that can only happen if the electron is pointing spin up.
NARRATOR: However, the universal quantum computer these guys are developing still requires a lot of work. Which is why, over in Canada, D-Wave has adopted a very different type of machine.
DR ERIC LADIZINSKY: The model that we chose is fundamentally more robust against environmental disturbance. It’s simpler to realise on realistic time frames for investors and all that to build something useful sooner.
NARRATOR: This quantum computer works in a very different way. The D-Wave uses electrical circuits with superconducting currents running through them which produce magnetic fields. The circuits behave like magnets and interact with each other.
DR ERIC LADIZINSKY: Finding the minimum energy state of a lot of interacting quantum magnets—the mathematical structure of that is very similar to a lot of really hard problems. We’ve said, ‘Let's build a physical system that finds the answer to a problem physically’. It’s not changing the problem into a bunch of mathematical equations and solving it with digital logic. It literally is asking what the best arrangement of these spins, these interacting spins. It just evolves to that arrangement if we do it right.
NARRATOR: But if it’s a quantum computer, why is it so huge, I hear you ask.
DR ERIC LADIZINSKY: That box is really to keep out electromagnetic radiation. It’s like wrapping your radio in aluminium foil—you won’t hear anything. So it’s a big Faraday cage or shield. So, for instance, the chip down here, you know, it has a cover on top of it, which is a radiation shield. This whole thing will be under vacuum, so there’s no air molecules bounding into it.
NARRATOR: D-Wave’s fast and furious approach has meant they’ve actually got a saleable prototype out there. The drawback is its limited application.
DR ERIC LADIZINSKY: It’s not a general purpose quantum computer. It’s application specific. If you're trying to minimise the risk of some financial portfolio, the mathematics are similar. If FedEx wants to find out, out of all the possible ways we could route our trucks, how do you do it to minimise fuel consumption. Those are all find the best of a vast number of possible solutions.
NARRATOR: These are all examples of optimisation problems. But, of course, there are many other types of problems out there too.
DR GRAHAM PHILLIPS: So, the race to build a universal quantum computer is still very much on, and these guys in Australia are frontrunners.
NARRATOR: When realised, the universal quantum computer will solve in seconds problems that a classical computer would’ve taken millennia to figure out.
PROFESSOR MICHELLE SIMMONS: We really want to build a kind of a universal, large-scale quantum computer with error correction that can solve all the quantum algorithms that we know exist out there.
NARRATOR: Solving known problems, like treating disease.
PROFESSOR ANDREW DZURAK: A very important one is to simulate the way that atoms and molecules are put together and connect. Actually designing new types of molecules, perhaps drugs for the pharmaceutical industry.
NARRATOR: And addressing problems we’re yet to even discover.
ASSOCIATE PROFESSOR ANDREA MORELLO: What we are making is such a completely different object that it’s really hard to even imagine what it would be good for. The only purpose of telling you now what we think it’s going to used for is so that in 10 years from now I can listen to myself saying, ‘Oh, why did I say that? It's so blatantly wrong’, you know?