Ultimately what I’d like to do is show you guys what a quantum computer processor really looks like. A lot of this material would be confusing if I just dumped it without context, even for experts, so I’m going to try to provide some perspective and background for some of the things I’ll be showing you later on about our machines. I’ll start by introducing the basic idea behind the whole field of quantum computation.
This basic idea is really pretty simple. But it’s still pretty deep. This basic idea is that
Information always has to be written on something.
Often it’s convenient to forget this basic fact and to view information as something abstract – a collection of ones and zeroes that lives in the ether somewhere. But no matter how hard you try to imagine otherwise, those ones and zeroes ultimately reside as physical states of something – marks of graphite on paper, directions of magnets in a hard drive, assemblies of chemicals in your brain, etc.
This basic idea has a big consequence: if information is always written on something, then that information has to behave according to the laws of physics of the stuff it’s written on. For example, if I write some words on a piece of paper, and that paper could travel back in time, then I could send information back in time. If the laws of physics say I can’t send pieces of paper back in time, then I can’t send that information back in time either (by the way, I don’t know if the laws of physics allows this or not, closed timelike curves notwithstanding).
This is relevant how? Well, computers have always stored information on devices that to a good approximation obey the laws of classical physics. This means that all that information has always had to behave according to those same laws. Conventional computer science implicitly assumes this, and up until recently all of the results, theorems, algorithms, etc. of computer science were based on the assumption that the machine embodying all of the information, and the calculations done using it, were entirely classical.
Nowadays we can build things that obey different laws of physics – namely those of quantum mechanics, which are generally more likely to be required to describe things that are small and cold. Quantum mechanics is a framework within which we can describe the properties and behavior of (probably) all matter and energy, and is conceptually quite different than conventional physics.
When scientists store information in quantum mechanical devices in the lab, what is seen is that this information does, in fact, obey the laws of quantum physics. In theory, what this means is that a large part of computer science, developed over the past century or so, needs to be examined in order to find out what difference it makes to computation if information is stored in a quantum system as opposed to a classical one.
This re-examination of computer science, for quantum computers, is still in its early days. There are already several algorithms developed for quantum computers that vastly outperform their classical counterparts, but the story is by no means complete.
Another way of putting the basic idea that “information is physical” is this:
What can be computed depends only on the laws of physics obeyed by the machine running the computation.