SIEMENS

Quantum physics is already a hundred years old, but it seems that its practical applications are only now becoming a reality.

Zeilinger: That's not true because all semiconductor technology is based on quantum physics, and lasers would also be unthinkable without it.

Certain phenomena addressed by your work, such as entanglement of photons, weren't even studied or applied until a few years ago. Why not?

Zeilinger: It's true that studies concerning photon entanglement didn't begin to emerge until the 1970s, although Erwin Schrödinger had described the phenomenon as early as 1935, a time when Albert Einstein also was examining the "spooky action at a distance," as he referred to it. The phenomenon was then neglected for several decades because scientists considered it a question for philosophers rather than for physicists. I wasn't taken seriously when I began looking into the question with some colleagues in the 1970s, and only about 30 people attended the first conference we held on the topic. Today, there's a conference every month attended by hundreds of researchers, more than I can even go to because of time constraints. So we are experiencing a revival of interest in the fundamental questions of quantum physics, and several entirely new applications have been emerging since the 1990s, including quantum cryptography.

What are the benefits of using quantum physics to help with data encryption?

Zeilinger: It makes encryption completely secure because you notice immediately if someone is trying to eavesdrop on the fiber-optic line that's used to exchange the keys. This isn't some kind of technological trick; it's a fundamental aspect of physics. In the project that we worked on with Austrian Research Centers, which Siemens was also involved in, we used entangled photons to transmit the keys. One photon always knows the state of the other, so if one is measured, say by a hacker, we can immediately see this in the measurement reading for the twin. We are now able to generate twin photons at a rate of ten million per second. We used around a dozen keys per second in the project with ARC, and that represents a completely new dimension compared to conventional procedures in which keys are exchanged on an unsecured line, and possibly used for years at a time.

Are there drawbacks to your procedure?

Zeilinger: Because you can't measure the photons without destroying the entanglement, you unfortunately can't reinforce them. Losses in the fiber-optic lines lead to a loss of photons, which is why the maximum bridging distance in the line has been limited so far to around 30 km. During wireless transmission experiments between the islands of Tenerife and La Palma, however, we have achieved a distance of 144 km. That's more than enough for the purpose of secure data transfer between government offices or banks in a given city. We also have to be careful to prevent side-channel attacks, which eavesdrop on the electromagnetic signals created in the transmitter and receiver when the keys are generated in the crypto-hardware system. But we've already taken appropriate measures to deal with that.

Are there any other applications for quantum physics that are ready for the market?

Zeilinger: Yes — generation of random numbers, for example, which are needed for online gambling systems, certain types of optimization algorithms, and for calculation of integrals in mathematical problems. Today such random numbers are generated with computers by setting a starting value for the program and then letting it run for a certain period of time. If the circumstances are the same, though, you'll end up with the same numbers, which is why these are called "pseudo-random" numbers. Someone working in a computer center who understands how a certain process works can use that knowledge to hit the jackpot at an online gambling site. We, on the other hand, generate true random numbers by shooting photons onto a semitransparent mirror and then measuring when the light particles either pass through or are reflected, which is a completely random process that can't be predicted. We've developed a market-ready generator that creates billions of random numbers per second, and we're already talking to companies that want to build and sell it. In general, however, quantum physics is currently in the same application stage that we saw with semiconductors and lasers when they were still in their infancy. By this I mean that the inventors initially didn't realize how many different things could be done with their inventions.

But scientists seem to have a very firm idea of what can be achieved with quantum computers, although they understand it will take a long time to do this.

Zeilinger: There's a big race going on with quantum computers today. When a quantum computer is actually built and can function properly, it will be in a class by itself, truly unprecedented. The exceptional feature of such a computer is that it will be able to process several operations simultaneously rather than in succession, because it will exploit the quantum mechanical overlapping of atoms. (Editor's note:see article "Catching Worms with Quanta")

What can be done with such tremendous computing power?

Zeilinger: One application would be Shore's algorithm for breaking down large prime numbers, which is a necessary component in the process of cracking an encrypted code. This would amount to the counterpart of quantum cryptography — but it would be useful for breaking only today's conventional codes, not those generated by quantum cryptography. Another application could involve a search for a name in an unsorted database, for example. When you use conventional algorithms here, you might end up having to search the entire database if you're unlucky, which would mean a million computing steps for a million entries. A quantum computer with just eight quantum bits, by contrast, could do the job in less than 4,000 steps.

Some of your colleagues doubt that there will ever be a quantum computer. They say that phenomena in the quantum world cannot simply be transferred to the macroscopic world.

Zeilinger: I'm absolutely certain we will eventually see quantum computers. There is no obstacle inherent to physics that would prevent this. According to Moore's Law, the number of transistors that can be placed on a chip doubles every 18 months. So, you could also say that the number of electrons needed to store a bit is reduced by half every 18 months. If you project that, you'll find that in 20 years only one atom will be required to store one bit — and with that we have the quantum computer. We have a lot of work to do, of course, because we still aren't able to control complex quantum systems. Nevertheless, it's only a question of time until we will do this, and some day every cell phone will contain such a quantum computer.