The catch? Quantum computers aren’t very useful yet. They currently require some high-grade scientific equipment to function, are expensive to build and maintain, and are only good at specific tasks. The bottom line is that quantum computers are almost at the point of being amazing scientific machines, but they may never actually speed up the delivery of cat pictures from Internet servers to your eyeballs.
Too long, didn’t read
Quantum computers are complicated, so if you’re just looking to get the idea without getting into the details, this is for you.
Conventional processors work by having lots of little parts that can be “flipped” to either a 0 or a 1 position. Superposition: a “Schrodinger’s cat” scenario: something can exist in multiple states until it is observed. For quantum computers this means it can store 0 and 1 simultaneously until it is called upon to be one or the other. Quantum entanglement: a property that lets quantum particles talk to each other – even over distances of many miles, any changes made to one particle will also affect the other. This lets quantum computers combine “superpositioned” chips to exponentially increase speed and storage space. Two bytes can store only one of the following: 0-0, 0-1, 1-0, or 1-1. Two qubits can store all of those. Qubits: Conventional computers use bits and bytes; quantum computers use qubits. These are the things that exist on the plane between 0 and 1, and they’re what everyone is trying to entangle and put on chips. Quantum computers aren’t very useful for day-to-day computing things, but they’re going to be insanely good at some very complex stuff.
Conventional Processors
Conventional processors, like the Intel or AMD chip in your computer, are essentially calculators that follow logical pathways – they get some data and a set of instructions telling them what to do (math, like adding/multiplying; logic, like AND/NOT). They perform the operation and send the result to be stored somewhere else. It’s that simple, an input/number goes in, and an output comes out; if it seems abstract, just imagine a black box that takes instructions and materials and spits out a product. If you have a 2.4Ghz processor, your computer is doing about 2.4 billion of these operations per second. The more numbers you can get out of your processor per second, the faster your programs will execute. On the hardware level, processors are composed of millions or even billions of transistors, which are essentially tiny little switches that are constantly being toggled (they don’t move, just change their electrical charge states) to represent one of two states: 0 or 1. These are arranged into logic gates, caches, and other fancy things on the chip, but the only thing we need to know is that transistors only have two possible states: they are always set to either 0 or 1, allowing one calculation to be done at a time. To summarize: conventional processors do billions of very simple operations very quickly using millions/billions of transistors arranged in certain patterns and set to either 0 or 1, depending on instructions.
Schrodinger’s Cat and Superposition
Instead of getting straight into the nuts and bolts, it’s best to start with some pretty out-there physics. (Don’t worry; there’s no math.) Schrodinger’s cat is one of the most famous examples of quantum physics, and it deals with the idea of “superposition.” It’s pretty simple: a scientist has a box with a cat inside. The cat has a 50% chance of dying. (No cats were harmed in the making of this illustration.) The scientist hasn’t opened the box, so he doesn’t know if the cat is alive or dead. From an objective standpoint, the cat must be either dead or alive, but from a quantum physics standpoint, both are true, at least until the box is opened. Why? Because (for our purposes, at least; there are lots of different ways to approach this) the scientist’s processing unit (his brain) doesn’t know what the answer is except that it could be either a live or a dead cat. In theory, the scientist has prepared for both possibilities, so when he opens the box, his brain receives the input (the cat is alive!) and produces the precalculated output (relief, presumably). This is superposition: the idea that something exists in multiple states until it is observed, measured, or otherwise acted upon. How does this apply to quantum computers? Just replace the scientist’s brain with a processor (metaphorically): it already knows the different possibilities (the instruction could either be for a 0 or a 1), and it is storing all possibilities at the same time. When it comes to output, though, it outputs a 0 or a 1, just like a normal processor. All possibilities can exist simultaneously, but only one output can emerge. It isn’t especially useful with just two numbers, but once you scale this up to the point where quantum computers can calculate billions of possibilities at a time, the potential starts to become obvious. As an analogy, imagine tossing a coin in the air. While it is flying, it is constantly rotating between heads and tails, effectively being heads, tails, and both heads and tails. That’s what a quantum computer’s processor is doing, and that’s why it can calculate pretty much every possible outcome instantly.
Quantum entanglement
Things start to get really interesting here. It turns out that quantum particles can exist in pairs and that each member of the pair is a mirror image of the other. This is “quantum entanglement.” If something happens to Particle 1, an opposite change will occur in Particle 2. Einstein called this “spooky action at a distance” because of how odd this property is. Military researchers are even experimenting with using it to replace radar – just fire one half of an entangled pair up into the sky and see what happens to its partner down here to figure out if it hit a plane. This is a bit mind-bending to get your head around, so suffice it to say that quantum computers can use entanglement to connect multiple “quantum transistors” or “qubits” to increase the level of complexity exponentially. A computer can look at the state of one qubit and then figure out what all the other ones are up to as well because they’re entangled.
Qubits
This is where the hardware comes in. Qubits are, like conventional computer bits and bytes, the most basic unit of quantum information storage. The big difference is that each qubit exists, in a sense, as both 0 and 1 simultaneously, which can be replicated on computer chips in a few different ways, from supercooled superconductors to lasers. The end goal is the same, though: get some kind of particle to exist in that weird quantum state where it is two things at once. For example, a tiny strip of supercooled metal can bounce electrons around with very little resistance, creating the potential for any state rather than holding the qubit in one state.
A good next step is to entangle the qubits, which essentially means you need to sync them all up to the same frequency so they can work together. This makes quantum computers a lot more powerful since getting the qubits entangled is what lets you have a whole chip of them working together. On its own a qubit is fairly impressive, but it doesn’t do anything too exciting. When it’s entangled with another qubit, though, it can store all possible values of both qubits combined: 0-0, 0-1, 1-0, 1-1, with 2^2 possibilities. If you entangle three qubits, you now have 2^3 possibilities (8). The world record chip as of June 2018 has 72 qubits which could, in theory, perform as many calculations in a second as a personal computer could in over a week. To make this a bit simpler: If you compare two conventional bits to two qubits, the most noticeable difference is that two bits can only be 0-0, 0-1, 1-0, or 1-1 — just one combination of binary results. Two qubits, however, can store all four of those at the same time, and since they grow exponentially, a few qubits goes a lot farther than a few bits. 3 entangled qubits can be 0-0-0, 0-0-1, 0-1-0, 0-1-1, 1-1-1, 1-1-0, 1-0-0, and 0-1-0, simultaneously – keep scaling that up one power at a time, and you end up with a computer that can store some very complex possibilities.
Coming soon (for a few specific things)
So this is a quantum computer: a machine that knows all the answers but only gives out the one that matches with the question. It’s a mind-bending machine, but it’s been built, and it’s getting bigger and better so quickly that it’s hard to keep up. You may be wondering when you’re going to get a tiny sub-arctic freezer filled with spooky science in your computer, and the answer is, unfortunately, not soon. That’s not to say it’ll never happen, but right now it can pretty much only function inside of a lab, and your five-year-old laptop can probably beat a quantum computer at most things. Quantum computers will be very good at a few things, though, like:
Breaking encryption: You don’t need to own Bitcoin to be worried about encryption breaking down. It’s what keeps pretty much everything on the Internet from being openly readable to anybody who wants to drop in and take a look. Your Wi-Fi? Encrypted. Credit card? Encrypted. Breaking RSA encryption is considered impossible with normal computers, but that’s just because they can’t guess fast enough. Quantum computers are amazing at guessing. Luckily, quantum entanglement seems like it might provide a new way to encrypt things. Searching huge amounts of data: Quantum computers can take a look at the data, store all the answers, and answer your question immediately. Say you have a random list of numbers, and you know that the number 193,201 occurs somewhere in it. A conventional computer has to cycle through all the numbers to find it, but a quantum computer knew where it was before you even asked. Modeling extremely complex scenarios: Chemical structures, physics problems, weather forecasting massively complex systems with lots of possible outcomes – that’s where quantum computing shines. Because it can exist in so many possible states at once, it can replicate the actual complexity of the variable-filled natural world (which is itself in a quantum state)
Quantum computers as they currently exist look like they’ll mostly be problem-solving machines, optimizing supply chains, powering artificial intelligence, predicting the weather, playing the stock market, etc. IBM, Intel, D-Wave, Google, and other companies are already producing versions of these machines and researching ways to make them more practical and usable. One significant hurdle, though, is that since qubits are built on calculating so many possibilities, quantum computers get things wrong sometimes. Researchers are working on fixing this, but it’s another reason why you probably won’t have a quantum computer replacing your much more mechanical (and therefore accurate) processor.
Conclusion: Confusion, but that’s okay
Take comfort in this: Most people have no idea how the stuff inside their computer works, and even the ones that do have an idea probably don’t understand everything about it. The best thing about specializing the way we do is that you don’t have to have a clue how your processor works in order to do amazing things with it, and the same will be true with quantum computers. The key difference is that while your Intel i7 is pretty neat, learning about it probably won’t make you question the very nature of reality.
