Quantum Computing: More Hype Than Reality? | by Alastair Williams | One Blue Planet | May, 2021

Amora R Jelo

Investors are getting excited, but quantum computing still has a long way to go before it is useful.

Alastair Williams
Photo by Michael Dziedzic on Unsplash

The age of commercial quantum computing is dawning — at least if you believe the marketing hype. Investors certainly seem to. The stock market debut of IonQ, a start-up focusing on quantum computers, is expected to attract a valuation of two billion dollars. That is a lot of money for a company that earned just $1 million last year.

The sky-high valuation suggests quantum computing has a rich future ahead of it. Forget that paltry million dollars of revenue — investors are expecting IonQ will start pulling in huge sums as the quantum market takes off. IonQ, they are hoping, will be the next Google or Apple. Are they right?

On the surface IonQ looks like a promising investment. Founded in 2015 by two quantum physicists, it has since attracted both funding and staff from some impressive companies and organisations. None of that, though, matters if quantum computing itself cannot live up to high expectations.

Quantum computers are not, as people often seem to think, a wholesale replacement of modern classical computers. They aren’t the next evolutionary step in the rapid growth of computing power; they are instead something quite alien, based on different physics and different logic.

Those differences make them more powerful than current computers in certain respects. They are capable of running thousands, even millions, of calculations in parallel. That makes them very good at solving certain kinds of problem, but doesn’t necessarily give them an advantage at everything.

Even where quantum computers do hold a hardware edge, they still need good software to make practical use of their power. That is proving a challenge. Unlike classical computers, which rely on well-established rules of logic, quantum computers work in strange and sometimes mysterious ways.

That means you cannot just take an algorithm that works on a classical computer and expect it to run faster on a quantum computer. First, it is unlikely to work at all — and even if it does, there’s no guarantee it will be faster. To see any improvement, the algorithm needs to be redesigned to take the special nature of quantum machines into account.

What’s worse, the problem itself may not be one that a quantum approach is better at solving. Some problems already have optimal solutions we can run perfectly well on classical computers. There is nothing a quantum computer can do to improve that.

Instead researchers need to identify problems that these computers can solve faster. New algorithms must be written, ones specially designed to take advantage of quantum hardware. This is tricky. Quantum computers rely on hard to understand phenomena and work in a probabilistic way, rather than the solid black and white way programmers are more familiar with.

What’s worse, the field still has a long way to advance. Quantum hardware is limited, expensive to build and hard to access. Even if researchers can invent an algorithm to solve a problem, they struggle to test it and demonstrate its superiority over a classical computer.

That’s to say, though quantum hardware is advancing, quantum software remains a long way behind. Demonstrations of quantum computers, even quantum supremacy, rarely show anything that might actually be useful in the real world. So far, then, quantum computing is more of academic interest rather than big business.

So why are investors so keen on IonQ? Part of the answer is hype, of course. Quantum computing sounds exciting. It is easy to build it up as a futuristic technology that will change the world. But equally, there is some substance behind the enthusiasm.

The quantum approach offers a different way to think about computation. The algorithms it does enable will allow us to tackle problems that even the most powerful supercomputers currently struggle with. These are problems that are interesting precisely because our modern computers are so bad at them.

One of these deals with simulation of the natural world. We live, at fundamental levels anyway, in a quantum world. Classical computers can model that, but only in a slow and ponderous way. Quantum computers, by contrast, are perfectly at home in a quantum world. Simulations that currently take months — models of drug molecules for example — could be done in mere hours with a quantum computer.

Parallel computation opens up other possibilities. Some problems are currently solved by lengthy brute force approaches — essentially trying every possible option to find the best one. Quantum computers can try all of these options at once, turning a problem that takes decades to solve now into a simple five-minute exercise.

That has benefits for optimisation problems — for example, trying to find the best delivery routes for a logistics firm. But it also threatens the cryptography the Internet relies on, potentially allowing anyone to crack even the toughest encryption systems. The consequences of that may be severe.

Good quantum hardware is at least a decade, and probably longer, away. The software to make use of it is even more distant — which means it will be a long time until we see any practical use of the technology.

There may, then, be a quantum future worthy of a billion dollars of investment. But don’t bet on this coming as rapidly as some suggest, and don’t expect it to achieve everything its biggest supporters are claiming. The world will be classical for a while yet.

Next Post

Changing lives by connecting all Americans to broadband internet

Jackson County, Kentucky, has one stop light in its 347 square miles—but also high-speed fiber optic internet service to rival any big city. In the coal country of eastern Kentucky, the 800-person town of McKee is the hub of a one-thousand-mile fiber-to-the-home network covering two of the nation’s poorest and […]