You Should Probably Start Caring About Quantum Computing Now

Since the words started entering the tech community vocabulary in the not-too-distant past, quantum computing has remained a specter. It's a change-everything technology of the future, shooting out research news in short bursts, but it's come together rather poorly in the public awareness. The big names that drive the tech narrative--Apple, Google, Microsoft, etc.--aren't so much driving quantum research and, hence, no narrative. A new special issue of Science suggests it's time for that to change: the quantum computing revolution is real, and it's getting close.

Let's recap the general quantum computing idea. In classic, present-day computing, information is carried by electrical charge representing either a "1" or "0." A bit. That's a very, very small unit of information. It's tiny enough that computing has been able to more or less faithfully adhere to Moore's law since its inception, doubling processor speed roughly every two years.

The charge-based universe has limits though and performance gains aren't what they used to be. It's become impossible to maintain that steady rate of improvement using single processor systems, and pushing computers farther has meant threaded and parallel processing. (How many cores does your computer have?) We're on a theoretically endless trajectory of stacking processors now, which could take on whole new dimensions in the future when that stacking becomes distributed across different computers in a network--a crucial realization of the cloud.

But there remains a physical, unavoidable limit in classical computing that is just the limit imposed by the size of electrons vs. the size of electrical components. ("Size" and electrons is a difficult idea, but let's just pretend for now it isn't.)  Sooner or later, you hit a wall where past it, you can no longer control electrons sufficiently for computing. This is where quantum computing comes in. Instead of using charge to transmit information, quantum information uses a different particle property: spin. Welcome to the world of spintronics. Spin is a highly unique property in that instead of just transmitting just either a 1 or a 0, it can do both at once in different combinations (see illustration). It's pretty weird, but potentially handy.

From Science:

Right now, researchers are realizing two big gains in quantum computing, both absolutely crucial in bringing the whole idea to reality. The problem is that handling qubits, the quantum bit of information, is pretty tough. To use them practically will require whole new algorithms and architecture, and we'll need ways of controlling the highly tempermental qubits without destroying them. Qubits have a habit of decohering easily. Basically, the slightest disturbance will cause a quantum system to fall out of whatever state it was in and become practically meaningless, with no hope of recovery. Imagine a mail carrier having to balance every message on a fingertip--and all of those messages are made of glass.

Or maybe messages made of ice would be more appropriate. Handling qubits at normal room temperatures has been impossible until now. In fact, systems have had to be held at near absolute zero to keep qubits intact. But research out this year from researchers at Stanford and IBM found that by using diamond-based materials, it'd be possible to run a tabletop system without compromising qubit integrity. It won't be the only idea for solving the temperature problem but, so far, it's out ahead.

The other major advance also has to do with coherence. You could say that qubits don't especially want to be qubits and, under normal conditions, will stop being weird and start acting like normal classical bits. Researchers over the past few years have demonstrated that it's possible to keep qubits together for several seconds with the help of good ol' silicon. Granted it's a highly pure form of silicon developed only recently, but it is still fundamentally the same stuff that computers rely on now.

What's more, research done at Princeton by Stephen Lyon and his team (published in 2012 in Nature Materials) demonstrated that by using this highly pure form of silicon, it's possible to control the spin state of billions of electrons. Working with large numbers of qubits has been on ongoing challenge of quantum computing research. Most work so far has involved just a few at once (as much as 84), whereas successful computing could require thousands or millions. Note that controlling the spins of billions of electrons is not the same thing as harnessing billions of electrons into a quantum computer, but it's something nonetheless.

We now return you to your awesomely powerful, amazingly compact classical computer, soon to be obsolete.

Reach this writer at michaelb@motherboard.tv or @everydayelk.