Separating quantum hype from quantum reality

Simon Benjamin is the co-founder of Quantum Motion and professor of quantum technologies at Oxford. Here he argues against a previous FT Alphaville article that said quantum computing was a classic bubble.

Quantum computing is a fast-growing, much-hyped industry. But suggestions that it’s purely hype — and when the bubble bursts we’ll be left with nothing of any value — is a misconception and a failure to understand where we are and where we’ll end up.

First a disclaimer: I’m far from an objective bystander. I’m the professor of quantum technologies in the Materials Department at Oxford, cofounder of the London-Oxford company Quantum Motion, and for two decades I’ve worked on how to build a quantum computer.

Many quantum computing companies exist today, but they generally aren’t making any money yet. But those companies are obviously now in R&D mode. For example, Psi Quantum, among the largest of the new players having raised over $665mm, does not engage commercially at all. It simply tells investors it will take time. And it will. I believe it may be the end of the decade before we have really impactful quantum computers.

In some areas the previous FT Alphaville article by Nikita Gourianov on the “quantum computing bubble” is both right and wrong (quantum pun intended). We know that many important things won’t go faster with a quantum computer. For example, the task of rendering graphics is made up of a vast number of individually easy calculations — going quantum won’t help.

And not every business will benefit from quantum computers, at least at first. The earliest impact will be in areas related to materials science (including energy materials), chemistry, or optimisation (possibly stretching to logistics/transport). Even in these sectors, businesses need only get involved if they want to be part of the enabling technology rather than a user. Others can relax despite calls to become ‘quantum ready’ — they won’t miss the quantum bus, because the bus is still being built.

But to suggest that there will never be high-value applications, and that quantum computers will never repay their R&D investment, is wrong. For evidence, Gourianov’s article looks to two areas: breaking codes, and accelerating discovery in chemistry and drug design.

It is well known that Shor’s algorithm gives an “exponential quantum advantage” — the strongest level of advantage where the practically-impossible task of breaking encryption suddenly becomes easy. The article objects that even so, there’s little commercial value because alternative codes will be adopted. We should certainly hope that’s true — I’d rather not see the world lose its ability to exchange data safely, since that is the enabler for all online finance and internet commerce, and essential to modern society. And how would it be ethical to sell access to a code breaker anyway?

Fortunately, the crypto community is indeed developing “quantum safe” codes. But for investors the significance of Shor’s algorithm was never commercial code-breaking. It’s that we can prove quantum computers will be able to do something that is practically impossible conventionally, thus establishing that they can be amazing, disruptive machines.

On chemistry, the critique points to a very recent preprint by over a dozen well-regarded authors. This paper considers whether there is yet evidence for exponential quantum advantage — Ie the strongest possible advantage — for a particular task: evaluating the “ground-state energy” of a molecule. The authors conclude that there isn’t such evidence yet.

But they say nothing about other levels of advantage nor indeed other tasks of great interest to chemists. For example, they note that “we cannot conclude anything about [quantum dynamics for chemical systems] based on this work”. Thus contrary to article, the paper inflicts no mortal wound on quantum computers as revolutionary tools in chemistry. It is just keeping the community honest.

Of course, literally hundreds of research papers identify and explore prospects for quantum advantage in areas ranging from optimisation (which is key to challenges in logistics and portfolio management), through to unlocking the dynamics of complex systems in the natural and technological worlds. Are all these ideas wrong, and ultimately unable to provide value? Almost certainly not.

So is there any elephant in the room for the nascent quantum computing industry to worry about? In fact there is. The issue is size.

Today’s prototype quantum machines are about the size of a wardrobe (or at worst, a full set of luxury bedroom furniture). But they don’t contain many of the qubits that are the raw processing unit of quantum computers: perhaps 100, usually fewer. We will need millions. And that’s sobering. Whichever of the leading approaches you consider — superconducting qubits, or ion traps, or pure-photonic — scaling up is likely to lead to a single quantum computer occupying the floor of a large building, if not the whole building. That’s just one quantum computer with one user at a time.

Building-sized quantum computers would still be impactful of course — perhaps comparable with today’s $36B HPC sector. But the sheer cost of such systems would limit their markets, and the benefits they can bring. And there aren’t many options for shrinking them. To me, the natural route is altering today’s silicon chips to host qubits instead of bits, but at the same minuscule scale. Certainly some solution to the size problem is needed if quantum computers are to reach their full revolutionary potential.

It’s fair to acknowledge that there are disconnects between the widely-held ideas of what quantum computers can be, and the reality of what they will be. But it is very wrong to assert that there is scarcely more to the field than hype. Progress toward quantum computers is real, and the path to commercially important machines is clear with obvious milestones. In that respect quantum computing is like any other disruptive technology.

Beyond all this, there is one other concern I’ve heard: that the effort to build quantum computers will fail due to as-yet undiscovered physics. Such a possibility seems remote, but it cannot be ruled out. However, uncovering a deeper reality behind quantum theory, which has withstood the scrutiny of thousands of experiments for a century, might be no less exciting than quantum computers themselves!