In Andrea Morello’s hands, the future of computing looks beautiful.
Standing in a light-filled laboratory at the University of New South Wales, Morello grips a small gold-plated circuit board in his fingertips. Slightly larger than a matchbox, its surface is punctured by a constellation of tiny holes and overlaid with white shapes like raindrops running down a window.
The tails of the raindrops converge near the middle of the golden board, where a tiny opening is waiting for Morello to insert a small piece of specially manipulated silicon.
This sliver of silicon is the same material used to build normal digital computers, but it has been altered at the atomic level.
By replacing selected silicon atoms with atoms of phosphorus, Morello and his colleague Andrew Dzurak have taken a step forward in a global race to build a computer using the weird laws that govern the physical world at the tiniest, quantum, scale.
The dream of building vastly more powerful computers by harnessing quantum properties has been around since the 1970s.
For decades, theorists have filled books and journals with discussion about what such computers could do. But turning those dreams into physical reality has proven a slow process.
Now, using a variety of different technologies, Morello’s team and many other scientists around the world are getting closer to building a functioning quantum computer.
There have really been some big strides toward the development of a quantum computer in the last two decades, says Dr Michael Biercuk, a quantum computing researcher at the University of Sydney and a chief investigator in the ARC Centre for Engineered Quantum Systems.
“We keep checking things off that people have said are impossible. Researchers have moved from proof-of-principle demonstrations to working on the engineering challenges we need to overcome to build something useful.”
How do quantum computers work?
The computer on your desk works by manipulating pieces of information known as binary digits, bits for short. Bits can only have two possible values – either 1 or 0 – normally represented by means of changes in electric current in a circuit.
However in the second half of the 20th century, scientists realised that they could add a twist to this scenario by using special properties of matter that applies when you get down to the sub-atomic scale.
The first of these properties is known as superposition. Put simply, superposition means that physical systems, such as electrons, exist in all their theoretically possible states at once. It’s only when you measure them that you get a result that corresponds to just one of the possible states.
Scientists realised that that if they could harness this kind of quantum system, each “quantum bit” of information, or qubit, could actually be both a 0 and 1 simultaneously. As you add more and more bits together, this superposition would allow you to exponentially increase the power of your quantum computer.
The whole idea of quantum computing hinges on the concept of ‘entanglement’, explains Dr Michelle Simmons, director of the Centre for Quantum Computation and Communication Technology at the University of New South Wales. “It means that if you change the state of one qubit it affects the other qubits that it is entangled with in the system,” she says. “It’s where the power of quantum computing comes from.”
“Entanglement in quantum computing plays the same role as heat in an engine or electricity in a light-bulb,” adds Andrew White, a quantum physicist from the University of Queensland. “It’s the underlying phenomenon you need to understand to build a quantum computer.”
Cracking codes, designing drugs
As theorists thought more about theoretical quantum computers, they came up with specific problems that they would be ideal for solving.
“The thing that really got the thing moving was in 1994 when a computer scientist called Peter Shor came up with a theoretical algorithm that could be run on a quantum computer if it existed,” says Dzurak.
Shor’s algorithm had the potential to solve a problem at the heart of the systems we use to keep our data secure, called public key encryption. This encryption system relies on the fact that conventional computers struggle to figure out the two large prime numbers that have been multiplied together to form another even more enormous number.
Shor figured out that a quantum computer would be great at solving this problem quickly, explains Dzurak. “All of a sudden one could see an application for quantum computers that was something that a conventional computer simply couldn’t do in any useful time.”
Not surprisingly, Shor’s realisation that quantum computers were ideal code-cracking machines generated interest from governments and the military. Their funding support provided an enormous boost to the field.
Since then, other potential uses have emerged. One example is the designing of the chemical molecules that are at the heart of drugs. Currently, this is a process of trial and error, which is one reason the development of new medicines is currently so costly and time-consuming.
“Ideally what you would like is to design them on a computer,” says Morello. “Some of these molecules aren’t that big, they may be only 20, 30 atoms, maybe 50 atoms. But this is completely impossible on a conventional computer.”
Beyond this, it is fair to say that the number of known situations where a quantum computer will out-perform your iPod is currently small, says Dzurak. “It’s a handful at the moment. But a couple of them happen to be quite important applications.”
Yet it seems likely that once we have quantum computers to play around with, the number of potential applications will increase dramatically. “In 10 or 20 years I don’t think the big impacts from quantum computers will be in cracking codes,” says White. “We’re finding category after category of problems that will be vastly easier if you had a working quantum computer.”
Ion traps and beyond
The first experimental demonstrations of qubits emerged in the 1990s using a technology that had been developed for atomic clocks, called ion traps. An ion is an atom (or molecule) that has a positive or negative charge. Scientists found that if they suspended such charged atoms in a vacuum using an electromagnetic field, they could use laser beams to control their internal energy levels and to measure them, allowing them to perform operations on basic quantum states.
Ion traps have led the field of quantum computing since the 1990s, partly because ions in a vacuum are well separated from their environment.
Earlier this year, Dr Michael Biercuk and colleagues from the US and South Africa used an ion trap to build a quantum computer with a layer of 300 beryllium ions with interacting spins acting as qubits.
“The system we have developed has the potential to perform calculations that would require a classical machine larger than the size of the known universe – and it does it all in a diameter of less than a millimetre,” says Biercuk.
The computer Biercuk’s group built is of a type known as a ‘quantum simulator’, which uses a well-controlled quantum device to mimic another system that is not understood.
“In our case, we are studying the interactions of spins in the field of quantum magnetism – a key problem that underlies new discoveries in materials science for energy, biology, and medicine,” says Biercuk.
Quantum dots and silicon
Ion traps are not the only systems people are exploring for quantum computers. Other approaches developed or conceived since the late 1990s include using superconductors, photons, diamonds, nuclear magnetic resonance on molecules in solutions and many more.
Rather than trapping an ion in a vacuum, Morelli and Dzurak’s team are inserting phosphorus atoms in chips of silicon.
“In a sense the silicon is like a vacuum because we make the silicon very pure so that the only kind of thing that’s active are the atoms that we deliberately put there,” says Dzurak. The spin of an electron orbiting the phosphorus serves as the qubit.
In 2010, the same group of researchers showed that they could measure the spin of that single phosphorous electron, which was controlling the flow of electrons in a nearby circuit.
Now, in research reported in today’s issue of Nature , they have demonstrated the ability to both read and write information on a single electron bound to one phosphorus atom embedded in silicon.
For Dzurak and Morello, the beauty of a silicon system lies in the fact that it’s a technology conventional computer manufacturers are comfortable with.
“That’s the thing about silicon quantum computing and why we’ve had so much interest and funding, because we’re using the technology that is the platform of a trillion dollar industry today. It will look exactly the same. You’ll look at it and it’ll look just like a computer chip.”
Biercuk says the latest research by the UNSW researchers is a major advance towards realising silicon-based quantum processing and takes us closer to the ideal of an integrated quantum computer.
“A major goal for the research community has been to realise the same quantum-coherent functionality afforded by atomic systems in a scalable, integrated platform. Silicon is a natural choice from this perspective, based on decades of research on large-scale integrated circuits for microprocessors and advanced digital electronics.”
“Nonetheless the whole community has a long way to go before a practically useful quantum computer is available,” he says.
As things stand, quantum computing is roughly at the place conventional computing was at in the 1950s, says Michelle Simmons, whose team at the University of New South Wales created the world’s first functioning single atom transistor.
“The first transistor was built in 1947 and the first integrated circuit came in 1960,” she says. “Lots was happening in the 13 years in between, and that’s where we are at now with quantum computing. We’re trying to integrate all the components in the one chip, so to speak.”
For a quantum computer to do some calculations that are beyond a conventional computer, researchers estimate it would need to have somewhere in the region of 30 or so qubits operating together. For more powerful operations, hundreds of thousands of qubits are needed. By this standard, Dzurak and Morello’s team, and most other research groups, have some way to go.
For now, it’s too early to say whether any of the various models of quantum computer might eventually win the race. Many researchers think some combination of different approaches might be most useful.
“We’re not quite at the point of selecting between different systems,” says University of Queensland physicist Andrew White. “We really don’t know what a real quantum computer will look like in the future.”
Originally Published by ABC Science Online 20 September 2012