“If you think you understand quantum physics, you don’t understand quantum physics.” That quote is attributed to the physicist Richard Feynman, but it’s unclear whether he actually said it. Here is a more reliable Feynman quote from a 1995 MIT publication: “I think I can safely say that nobody understands quantum mechanics.”
Now that we’ve gotten that out of the way, let’s see if there’s anything that we do know. Quantum mechanics is weird. Those tiny particles at the quantum level just don’t behave as expected. Things are different there.
Crazy things are happening in the quantum universe. There’s the intrinsic randomness, the uncertainty, the entanglement. It all seems a bit much.
We know now that atoms and subatomic particles act as if they’re connected. Einstein called quantum entanglement “spooky action at a distance.” Imagine two objects that are physically apart but they behave the same way, they have the same properties, and they act as one. Now imagine that those two objects are 100,000 light years apart. Weird indeed.
There’s more. The uncertainty principle in quantum mechanics says that certain properties of particles just cannot be known. Add to that the problem of decoherence, which has something to do with the collapse of wave function. And versions of the double-slit experiment seem to suggest that one quantum object can be in two places at the same time, that observation changes the nature of subatomic particles, or that electrons appear to have traveled back in time.
Now you see why building a quantum computer can be such a challenge. But that’s not stopping people from trying. (For more on quantum computing, see Why Quantum Computing May Be the Next Turn on the Big Data Highway.)
The Making of a Quantum Bit
The problem with uncertainty is that it makes computation difficult. The target is always moving. And even if you develop some mathematical system, how do you correct for errors? And you thought binary was hard.
“A qubit is a quantum mechanical system that, under some suitable circumstances, can be treated as having only two quantum levels,” says Professor Andrea Morello of the University of New South Wales in Australia. “And once you have that, you can use it to encode quantum information.”
Easier said than done. Current quantum computers are not very powerful yet. They’re still trying to get the building blocks right.
A quantum bit, also known as a qubit, has exponentially more potential than the classical bit in binary digital computing. An elementary particle can be in multiple states simultaneously, a quality known as superposition. Whereas a classical bit can be in either one of two states (one or zero), a qubit can be in both of those positions at the same time.
Think of a coin. It has two sides: heads or tails. A coin is binary. But imagine that you flip the coin into the air and it keeps flipping indefinitely. While it’s flipping, is it heads or is it tails? What will it be if it should ever land? How can you quantify the flipping coin? That’s a feeble attempt at illustrating superposition.
So how do you make a qubit? Well, if quantum physicists don’t understand quantum mechanics, then we could hardly manage an adequate explanation here. Let’s settle for a shortlist of technologies being tested to create qubits:
- Superconducting circuits
- Spin qubits
- Ion traps
- Photonic circuits
- Topological braids
The most popular of these are the first two. The others are subjects of university research. In the first technique, superconductors are supercooled to eliminate electromagnetic interference. But coherence times are relatively short and things break down. Professor Morello is working on the spin technique. Quantum particles have electrical charge, just as magnets do. By sending microwave pulses, he is able to get an electron to spin up rather than down, thereby creating a single-electron transistor.
Then there remains the matter of fault tolerance and error correction. Researchers at the University of California, Santa Barbara have managed to reach 99.4 percent fidelity with their qubit gates. They have achieved 99.9 percent gate fidelity at the University of Oxford. So are we there yet?
How Close Are We?
Edwin Cartlidge asks this question in an October 2016 article for Optics & Photonics News. A warning from ETSI in 2015 that organizations should switch to “quantum safe” encryption techniques should tell you that something is on the horizon.
Google, Microsoft, Intel and IBM are all in the game. One of the thresholds Google is pursuing is something they have termed “quantum supremacy.” It is used to describe that point at which a quantum computer does something that a classical computer can’t.
IBM plans to roll out a “universal” quantum computer in 2017, according to David Castelvecchi in Scientific American. Dubbed “IBM Q,” it will be a cloud-based service available over the internet for a fee. You can get a taste of what they’re working on by trying out their Quantum Experience, now available online. But Castelvecchi says that none of these efforts are more powerful than conventional computers – yet. The supremacy of quantum has not yet been established.
As Techopedia reported in 2013, Google has plenty of applications for a mature quantum computer, once developed. Microsoft is working on topological quantum computing. Several startups are ramping up, and plenty of work is being done in the field. But some experts warn that the dish may not be fully cooked yet. “I am not making any press releases about the future,” says Rainer Blatt at the University of Innsbruck in Austria. And physicist David Wineland says, “I’m optimistic in the long term, but what ‘long term’ means, I don’t know.” (See 5 Cool Things Google's Quantum Computer Could Do.)
Even when quantum computing supremacy is achieved, don’t look for it to replace your laptop anytime soon. Quantum computers, like their binary counterparts in the early days, may just be specialized devices dedicated to specific purposes. One of the most commonsense uses would be to have a quantum computer simulate quantum mechanics. Aside from intensive computer operations like weather forecasting, usage of quantum computing may be centralized and limited to the cloud. Of course, that may be the perfect place for it.
Professor Morello clearly identified the primary challenge of quantum computing. Before you can begin to encode information, you have to be able to establish two discrete quantum levels with the qubit. Once achieved, quantum computing “gives you access to an exponentially larger computation space” than a classical computer. A quantum computer, for instance, with 300 qubits (N qubits = 2N classical bits) would be able to process more bits of information than there are particles in the universe.
That’s a lot of bits. But getting from here to there will take some doing.