It's difficult but you don't have to be the Canadian prime minister to get it
Quantum computing is in the limelight this week after Canadian prime minister Justin Trudeau explained what it is during a press conference at the Perimeter Institute in Waterloo, Ontario. We already know he’s awesome at everything from dancing to yoga, so why shouldn’t he be a quantum computing expert too? But some people have suggested that he staged the entire thing and planted the reporter so he could look smart. We think he was being legit because his explanation was less than perfect.
Trudeau explained that normal computers store information as 0s and 1s, meaning there are only two possible values for information. He said that because of uncertainty in quantum physics, the quantum computers will use many more values and the increased complexity makes them faster and smaller. This isn’t entirely correct but isn’t totally wrong either. The individual bits of information in quantum computers don’t contain more values. Also, the attraction of quantum computers isn’t that they’ll be super fast or smaller versions of normal computers.
It’s hard to explain quantum computers at all let alone in mere seconds so Trudeau should be praised for giving a quick answer that wasn’t terrible. It’s true that quantum computers could perform very complex calculations and there is uncertainty involved but it’s not about creating faster and smaller computers; it’s about doing entirely new things altogether.
We’re no quantum mechanics experts but we’re major nerds so let’s try our own explanation.
To understand quantum computers, you need to understand a bit of quantum mechanics. Richard Feynman supposedly said, “if you think you understand quantum mechanics, you don’t understand quantum mechanics.” As you can imagine, it’s really difficult to explain quantum computers in a way that makes sense to people who are used to traditional computers, which is all of us. You have to think differently. At our size, physics seems to work a certain way. If you drop a ball it falls to the ground. It doesn’t turn into two balls, or suddenly exist in two places at once, or remain both dropped and not dropped. Physics works a certain way and so do our computers.
At a smaller level, the same laws of physics are difficult for us to comprehend. For example, a particle can simultaneously exist in two opposing states at once until it is observed. Imagine if that’s how it worked at our scale. Imagine if lights weren’t simply on or off but were both until you actually checked. It sounds nonsensical but that’s just because we’ve evolved to understand how things work at this scale.
Another example is entanglement, which we’ve covered before. If two particles become entangled, their states are linked and instantly reflect each other regardless of the distance between them. If you observe one particle, which forces it to settle on a specific state (like our light being either on or off), its entangled partner immediately settles on the opposite. In theory you could put a particle on one side of the galaxy and its entangled partner on the other side and both would immediately settle on their states when one is observed. No signal is sent between the particles as there’s no delay between the particles. It’s just as if they’re still in the same space as far as the universe is concerned. Quantum mechanics is weird on stilts.
All this weirdness isn’t going to make traditional computers faster or smaller than they are today. That’s not the point. We’re not going to get amazing videogames because of quantum computing. Instead, we could make computers that can do things no normal computer ever could. That’s the appeal. They’re not merely faster or smaller; they’re fundamentally different beasts because they work in fundamentally different ways.
A bit of information in computing uses 0s and 1s. Trudeau got that right. A piece of wire either has electricity going through it or not. The idea that a quantum computer’s bits are more complex isn’t really true and isn’t what’s fascinating and promising about the new technology. A bit in quantum computing is called a qubit and it also gives you a 0 or 1. There’s no added level of information complexity from a single qubit of information. Confusingly, we do get more complexity using qubits once you start using a bunch of them together.
In quantum computing, the state of a qubit (let’s just call it 0 or 1) isn’t set until you check. Before the value is given, it’s both a 0 and 1. Isn’t quantum weirdness just marvellous? A quantum computer does a bunch of calculations and then measures the value of the qubits (to see if they give a 0 or 1). Bizarrely, some of the qubits will only have a specific value at this point, right at the very end of the calculation, when the computer checks. This means that some qubits were simultaneously a 0 and 1 at the same time during the calculation itself. Whoa.
To summarise: a bit can store either a 0 or a 1 but a qubit can store either value or even both at the same time. That’s the game-changer: storing multiple states at the same time.
Qubits wouldn’t be useful for most calculations. Could you do maths when some of the numbers are two different values at the same time? A normal computer needs to know what it’s dealing with and this is why quantum computers aren’t merely super versions of what we already have. What makes quantum computing special is that there are some calculations that normal computers can’t do but quantum computers theoretically could. Quantum computers won’t replace traditional computers; they’re for different tasks.
When Trudeau mentioned qubits allowing increased complexity he was right; it just wasn’t because of extra complexity in the qubits themselves. Because there is uncertainty over whether a certain qubit is a 0 or 1 during parts of the calculations, there are some very complex calculations that they should be able to do that normal computers can’t. Despite qubits only having a value of 0 or 1 when observed, quantum computers become exponentially more powerful the more qubits you add. So Trudeau is bang on the money about more complexity; it just doesn’t come from qubits having extra states.
If a traditional computer had three bits of information they might give us values of 0, 1, and 1 for example. One bit equals one number so we can describe this system using just 3 numbers: 0,1, and 1 obviously. The confusing nature of the entangled particles means that multiple qubits can be represented by numbers in a totally different way. Instead of using 3 numbers to represent 3 bits, it would take 8 numbers to describe 3 qubits because it takes 2ⁿ numbers where n is the number of qubits. If the quantum computer has 10 qubits instead of 3, suddenly you need 1024 numbers to describe it. If we build a 20-qubit computer, it would take over 1 million numbers to describe it. A single qubit doesn’t have more complexity than a regular bit but their very behaviour in a quantum computer means the information required to describe the computer doubles with every additional qubit.
Remember that physics seems kind of predictable to us? The ball drops when we drop it. The world just works the way we expect it to. Our computers feel the same way. We can teach a computer what happens when a ball drops. We can get it to calculate, model, and simulate the dropping of a ball. But quantum shenanigans? We struggle to comprehend them and our computers fail to as well.
Think about how it takes 20 numbers to describe 20 bits in a normal computer but over a million numbers to describe 20 qubits. The more information you want to store in a traditional computer, the more binary numbers you need, which means you need more transistors. If you want to do a calculation that uses numbers bigger than the number of atoms in the entire universe, it’s going to take an impossibly large computer a disgustingly long time. In fact, some calculations are practically impossible with today’s computers.
We’re having to make chips smaller in order to make them more powerful. Moore’s Law is already starting to fail and soon physics will become an obstacle and we won’t be able to go smaller. The appeal of quantum computers isn’t that they’ll do regular computing but let us build even smaller; it’s that quantum computers will be relatively small considering the calculations they could do.
A traditional computer could not do the quantum computations scientists want to do with quantum computers. You could not program a normal computer to run a simulated quantum computer. The beauty of quantum computing and its application is in doing the exact opposite: using a few particles and quantum mechanics to solve absurdly complex computational problems. If you wanted to play with numbers that are bigger than the number of atoms in the universe, you would only need a few hundred qubits. Quantum computers are still very difficult to make but last year D-Wave Systems claimed to have broken the 1000-qubit barrier.
For an example of a calculation quantum computers could do, take factorisation. Give your traditional computer a specific known number and ask it to find two unknown prime numbers that give the known number when multiplied. Your computer will scream “please, no more” and go to sleep. This is known as an “intractable problem” as normal computers can’t calculate large prime factors quickly, but quantum algorithms could calculate the prime factors of very large numbers very quickly indeed. When you give your credit card details online, the website’s encryption works because computers can’t calculate large prime factors. The whole reason it works for security is that nobody has a computer that could deal with the computation problem. In theory, a quantum computer could break through these encryption methods easily, which shows just how fundamentally different the technology is.
Computers can help explain the world around us. Think of the scientific models used to explain complex systems like climate change or biodiversity. But as much as we like to think that quantum mechanics is some alien concept, it’s just as real as the physics that we comprehend every day. If we want to understand reality and how the universe works at a fundamental level, we need to understand quantum mechanics too. The very pursuit of building quantum computers requires us to learn more about quantum physics. The development and use of quantum computers should help us better understand the universe in a way traditional computers never could.
That’s not a feature you’ll see on the next MacBook.
Main image © D-Wave Systems, Inc.