The field of quantum mechanics still seems as mysterious as ever, even 100 years after Niels Bohr first proposed his quantum model of the atom.
Quantum theory has given us waves that are also particles; particles that are somehow “entangled” with one another; and particles that are in “superpositions,” both here and there, or up and down.
But these same quantum properties could also hold the key to powerful new computers. With a working quantum computer, hard problems could suddenly become easy – problems like factoring extremely large numbers, or simulating complex molecular interactions.
Our everyday computers using the laws of classical physics work with bits of information that are either on or off. On the other hand quantum bits, or “qubits,” can be in superpositions that represent both on and off at the same time, creating tremendous potential computing power. A quantum computer of just a few hundred qubits could quickly solve some problems that a classical supercomputer would still be working on when the Sun had burned itself out.
But there’s a lot to be done. Practical devices have to be developed, and algorithms that exploit the quantum effects of superposition, entanglement, and quantum interference have to be invented. At a recent meeting of CIFAR’s program in Quantum Information Science in Sherbrooke, REACH sat down with Program Director Raymond Laflamme (University of Waterloo) and Senior Fellows John Watrous (University of Waterloo) and Barry Sanders (University of Calgary) to discuss the future of quantum computing.
Why are we interested in quantum computing in the first place?
Laflamme: Quantum computing allows us to do things that we cannot do with classical computers. Today’s computers use the rules of classical physics to manipulate information. As you go down to the atomic scale, the laws of physics change. What we have found is that the laws of quantum mechanics seem to help us manipulate information in ways that we can’t with classical computers. It’s not only an incremental change. There’s a discontinuity there.
Watrous: We have examples of problems for which quantum computers are apparently better than classical computers, according to theoretical models. According to these models one can, in principle, build a quantum computer that will factor numbers efficiently. And it’s not known how to do this classically. This and other examples give us reason to think that quantum computers have a lot of potential.
We’re talking about problems that if you try to solve them with a classical computer, running the fastest algorithms that we currently have, it would take more time than it will take for the sun to engulf the earth. Some problems are simply out of reach of classical computers.
Will these be specialized devices created for special problems? Or is there going to be a kind of general computing device?
Sanders: We’re not aiming to invent necessarily a new computer. If we can get things to work, then I think of it like an add-on chip. So we have our computers, and instead of some turbo device we get an add-on chip that will enable us to make some hard problems easy.
Quantum states are seen as tremendously fragile things that are always on the verge of collapse. But you’re talking about building devices that let you create these states and manipulate them.
Sanders: Yeah, that was one of the things about quantum mechanics, the fact that you can ignore it for almost everything in the world. That it might matter if you want to understand atoms, molecules, nuclei, but we could ignore it at the higher level. But this is where we’re pushing the boundaries. We can engineer systems so that quantum mechanics does matter on a larger scale, and we can exploit it.
What kind of a timeline are we looking at for working quantum computers?
Watrous: We’re coming up on 2014, which is the 20th anniversary of [Peter] Shor’s algorithm, which shows how a quantum computer could be used to efficiently factor numbers. And this was a question that people were asking then – how long is it going to be before you can build a quantum computer? And people always used to say, well, 20 years.
Sanders: It’s always 20 years.
Watrous: Right. But we’ve learned an enormous amount about the nature of the problem. We’re up against a pretty big challenge, but we’re optimistic.
Laflamme: If you think of a large-scale quantum computer, the timeline is still fuzzy. But today we have small devices where quantum effects are important, which can be used practically today. Advances in quantum metrology and quantum sensing have given us devices that allow us to create images of a single electron, for instance, and could someday be used for things like creating detailed images of proteins or other molecules of interest.
Sanders: I agree with Ray, that along the way there are all these benefits that come along. There’s the quantum metrology stuff, there’s the side benefits to nanotechnology, etc.
Where does Canada fit into all this?
Laflamme: We have people who are really, really strong. Something which is unique to the CIFAR program is the combination of physics and computer science. There’s some great stuff that has come out of the program because of that.
What are the big questions you’re interested in?
Watrous: I’m interested in the mathematics of quantum information. There are many unanswered questions, questions about entanglement, for example. Entanglement is very poorly understood in large systems. The types of correlations that can arise from entangled states, for example, is a huge mystery, and there are mathematical challenges to try to understand these things.
I’m also interested in questions related to computational complexity, which tries to understand the limitations on the power of quantum computers.
Barry, what about you?
Sanders: I’m interested in [physicist Richard] Feynman’s original question that motivated the idea of quantum computing, the question of whether nature can be simulated. You know the movie The Matrix? If you haven’t seen it I’m sorry if I’m giving anything away…
Watrous: It’s been around a long time. You don’t have to worry about spoilers.
Sanders: So then the idea is, do we live in the world or do we live in the matrix? And suppose there is a matrix – is the computer that runs it classical or quantum?
But the serious question is whether nature is simulatable. This is the question that really drives me, is whether the universe as we know it is equivalent in some sense to quantum computing or not. And then all the technology and stuff is what I need to do to earn my money to be able to think about it.
How about you, Ray?
Laflamme: So I’m interested in knowing can we, and how can we, control quantum systems. If you look at the history of humankind and you lookat technological evolution, from fire 20,000 years ago, or steam 200 years ago, or electricity a hundred years ago, you can see a pattern that starts with people being curious about something.
We’re curious about some things, and then by kind of pushing our brain we learn how to understand these phenomena of nature and kind of see how they work. And once we understand them, then we have a path of how to control them.
Quantum mechanics is at least a hundred years old. About that time we started to realize that there was this new phenomenon of nature. It took time to build a theory, and we’ve learned a little bit about controlling some small part of the quantum world, but we haven’t learned how to control the superpositions of many states, and this is the essence of quantum computing and quantum information science. Controlling is the important part. That’s why I do what I do.
Watrous: You have a much more romantic answer than me. I just want to prove theorems.
Laflamme: And that’s the fun part of interacting with you.
Watrous: Well, you are among the people who have proved one of the foundational theorems of quantum computing, which is the threshold theorem. I mean this is a fundamentally important result, which helps us to have confidence or at least hope in being able to build a quantum computer.
Laflamme: Yeah, so, I had help.