Billions of dollars are being spent on developing “quantum computers” that exploit some of the weirdest and most counter-intuitive phenomena in all of physics. Some companies have already sold commercial products for as much $10 million a pop.
There’s just one problem: No one is 100% certain that these things actually work.
A quantum computer is so different from a conventional computer that even calling it a “computer” is more misleading than descriptive.
Virtually every computer built since the field was invented in the late 1940s works by reducing everything — words, sounds, images, equations — to a series of 1s and 0s, known as binary code. Here’s what my digitized first name looks like: 01001100 01100101 01100101.
Those 1s and 0s are then manipulated by transistorized switches. They do nothing more than switch back and forth between 1 and 0 but they do it very fast. The little switches are called “bits,” and if you put enough of them together in logic circuits, they do all the things you’re familiar with when using your personal computer, smartphone, and nearly everything else in the world that runs on electricity.
Quantum computers are different. Instead of bits they have “qubits” (quantum bits), and instead of being either 1 or 0, they can be both at the same time, or anything in between. If you didn’t understand that last part, don’t worry about it: Nobody does. But it’s true nonetheless, borne out not just by theory but experiment.
These simultaneous states are found everywhere in the quantum world. As an example, remember “Your Friend the Atom” from third grade, with that precious cartoon of point-like electrons whirling around a smiling nucleus? Cute, but wrong. Electrons actually surround the nucleus in a “probability cloud.” They’re everywhere and nowhere at once, with only a likelihood of being in some locations rather than others at any particular time.
Those likelihoods are defined by a wave equation. When something happens that triggers the electrons to occupy definitive positions, the wave function “collapses” to a set of specific points, and that’s where we find the electrons. And it’s that collapse that is central to the quantum computer, which might allow us to solve certain problems that a conventional computer can’t.
Suppose a traveling salesman has to visit three cities, and you’re asked to provide a route that minimizes his driving time. Pretty easy problem; all you have to do is look at a map and the answer is obvious.
How about five cities? A little tougher, but manageable. Twenty cities? Forget it. Not only are you unlikely to figure it out, if you do, there’s no way to prove you’re right unless you try every possible routing. That’s because there’s no mathematical formula that will give you the answer. And good luck with “brute forcing” your way through all the possibilities, even with a powerful computer: There are 2,432,902,008,176,640,000 (2.4 million trillion) of them.
A quantum computer, on the other hand, would be ideally suited for this kind of problem. If a wave-function-like equation describing the situation is devised, and a series of qubits is configured to reflect that equation, then all of those qubits collapsing to definitive states should provide the fully optimized driving route.
The goal, of course, is not to help out traveling salesmen. A quantum computer by its very nature should be able to mimic processes in the quantum world and yield extraordinary insight into the nature of the physical universe. It can also provide practical benefits, like enabling unbreakable encryption in the financial, medical, and defense arenas.
But how do we know when a so-called quantum computer is actually doing what it’s supposed to be doing and not “cheating” somehow, as some people in the scientific community suspect?
The answer is simple: Give the machine a problem that can’t possibly be solved in reasonable time by the most powerful conventional computer in the world. If it succeeds, it will have achieved “quantum supremacy” and we’re off and running to the future.
It used to be thought that a quantum computer containing 50 qubits could meet that threshold, but advances in conventional computing are proceeding very rapidly, and that number keeps rising. Two years ago IBM demonstrated a computer that could simulate a 56-bit quantum computer.
The most powerful conventional computer in the world right now is the Summit machine, designed by IBM for use at the Oak Ridge National Laboratory. It is capable of 200 million billion special instructions called “floating point operations” per second. For our 20-city traveling salesman problem, it would take Summit close to three hours to find the answer, using brute force to test every possible routing.
A quantum computer containing 80 qubits could do it in a fraction of a second, without testing any routes. Add two more cities, and Summit would take four months to crack it. The 80-qubit machine would still knock it off in under a second.
There’s no quantum computer with 80 qubits yet. The record so far is 72, held by Google’s Bristlecone chip. Adding qubits is fiendishly difficult; for one thing, they’re “noisy,” with error rates many billions of times higher than conventional computing chips. One proposed approach to addressing this problem, which is nowhere even close to being implemented yet, is to use thousands of specialized, error-correcting qubits surrounding each primary qubit to get those rates down to something manageable. Another is to determine whether there are “fault tolerant” ways of getting useful results out of these machines despite the errors.
As you might imagine, all of the foregoing is vastly simplified and therefore even regrettably misleading in spots. Tests of quantum supremacy are not going to involve trivial traveling salesman problems; the differences between quantum and conventional computers are so profound that entirely new types of tests have been configured to enable meaningful comparison.
Leading industry prognosticators have indicated in recent weeks that quantum supremacy stands a good chance of being achieved before the end of 2019, most likely by Google. While other groups are using such exotic technologies as ion trapping and “topological qubits,” Google is using superconducting circuitry, which is well understood and advancing rapidly. Progress following that milestone is likely to come at a brisk pace, but what the industry really needs is a major, fundamental breakthrough so compelling that all the players coalesce around a single technological direction.
Watch this space…
Lee Gruenfeld was Vice President of Strategic Initiatives for Support.com, Senior Vice President and General Manager of a SaaS division he created for a technology company in Las Vegas, national head of professional services for computing pioneer Tymshare, and a Partner in the management consulting practice of Deloitte in New York and Los Angeles. Lee is also a writer and photographer for the official website of IRONMAN, and the award-winning author of fifteen critically-acclaimed, best-selling works of fiction and non-fiction. For more of his reports — Click Here Now.
© 2021 Newsmax. All rights reserved.