Quantum computing

Imagine, for a moment, that the promise of powerful, super-fast quantum computers has materialized. In the beginning at least, there will only be a few of them, housed in special facilities.

Users who want to harness their quantum capabilities will need to send data to a remote location, allow the computer to do its magic and send back the results. Quantum physicists have now shown that there's a way to do this that's absolutely securely - meaning the remote quantum computer will never understand the true data even while it is manipulating it.

Though other researchers have described the theory behind such a "blind quantum computing" protocol a few years ago, a group of scientists in Vienna have now become the first to demonstrate that this works. Their results were published last month in the journal, Science.

Classical vs. Quantum

"Here is our laser," explains Stephanie Barz, a doctoral student at the Vienna Center for Quantum Science and Technology, showing off their experimental quantum computer that uses photons to hold and manipulate data. "Then we have crystals, we have optical elements like wave plates, we have beam splitters, optical fibers."

It looks nothing like an ordinary computer, more like a pinball machine - dissected into a maze of crystal cubes and optical lenses.

Using this setup, Barz led a group of scientists to demonstrate the possibility of blind quantum computing, which is one more step on the road to full-fledged quantum computing. Given the technology available right now, scientists have only been able to build very basic quantum computers that are neither stable nor powerful.

Since the early 1980s, computer scientists have been working on quantum computers, which promise to dramatically increase the power of computational power by taking advantage of properties of quantum physics, which allow data to be encoded into far more states than normal binary logic allows.

All "classical" computers - including the one that you're reading this on - can be fundamentally reduced into a series of zeros and ones, where those values are the only possibilities. But in the quantum world, a "quantum bit," or qubit can be any number of values.

"Quantum mechanics is based on numbers called amplitudes, which can be positive or negative," explains Scott Aaronson, associate professor of electrical engineering and computer science at the Massachusetts Institute of Technology, in the United States.

Aaronson notes that the different ways in which an event could happen is each assigned a particular amplitude. Take the event of snow tomorrow: the actual probability that it snows tomorrow is simply a function of all the amplitudes for all the possible conditions under which it could snow tomorrow.

"That's a very weird picture of the world," he says. "We're not born knowing that the world is this list of amplitudes.  But once you swallow that, everything else can be explained in a few sentences."

"A classical bit would be the North and South pole of a sphere," he explains. "A quantum bit can take all points on the sphere as a state. So it's many, many states. And that's what makes a QC so much more powerful than a classical computer."

Encryption

This is where Barz and her group come in - to imagine the future of quantum computing.

"So all these challenges in realizing quantum computers, this might lead to the conclusion that in the future only a few computing centers will be able to operate those devices," she says.

In other words, scientists will have to send their valuable data to a quantum computer on a remote server. Blind quantum computing is simply the absolutely secure way for users to do this.

The blind qubits are created using the configuration of crystals and optics . The remote quantum computer receives these qubits and a set of calculation instructions, but they are both encrypted such that the computer can perform the calculation without actually knows what kind of calculation it is performing.

"The server gets results like zeroes and ones, but for the server, these results appear totally random," Barz explains. "The client interprets the results and knows the output of the computation."

A popular misconception, however, is that quantum computers will be more powerful than ordinary computers in every respect. 

"It is not a panacea that lets us solve every type of problem faster," he says. "But for certain specific problems like code-breaking applications or like quantum simulation, it really does offer the promise of a very dramatic type of speedup."

In other words, for sending e-mails or playing Angry Birds, we are probably better off on our normal computers.

Step by step

This particular demonstration of blind quantum computing used a total of four qubits, still very small. Barz's group is considering on a version that uses more qubits and is capable of more complex calculations.

But, to take a step back - a stable, powerful quantum computer doesn't yet exist, nor do we have any idea when it actually will.

Still, Aaronson says he's fairly impressed with the results coming out of Austria.

"The results of these sorts of experiments are never surprising once you know the theory. Yet actually doing the experiment takes a lot of craftsmanship," he says. "It's very hard to get quantum mechanics experiments to work right."

Author: Sruthi Pinnameneni, Vienna
Editor: Cyrus Farivar