Enterprise Explains: The race for quantum computing and why that matters: At the tail end of last year we noted how Chinese researchers had declared reaching “quantum supremacy“ with their experimental ’Jiuzhang’ machine at the University of Science and Technology.

Why should I care? We wouldn’t be that surprised if that was the response of a sizable chunk of our readership. After all, the number of people who have an understanding of quantum mechanics let alone understand the significance of using photons over superconductors is probably quite low. But esoteric subject matter aside, the developments in China may come to fundamentally change how we understand the world around us, so hear us out.

First things first: Quantum computers make that brand new MacBook Pro sitting on your desk look proto-neanderthal. Seriously, even the world’s most powerful supercomputers are no match for what their still-primitive quantum cousins can achieve. The system developed in China processed in mere minutes a calculation that would take a supercomputer 2.5 bn years to figure out.

How does it actually work? Computers talk in binary. That is, they process information through a series of 0s and 1s, also known as bits. This classical binary system has underpinned the history of modern computing, from the first, most basic machine produced in the 1930s to the world’s most advanced supercomputers being built in 2021. Quantum computing changes all of this. While standard computers are stuck communicating in either 1s or 0s, quantum theory uses what computer scientists call qubits: units that can be 1, 0 or both at the same time. The idea that a particle can be in two places at once, or in this case be in two states at once, is known as a superposition and is a foundational concept of quantum mechanics.

Why is this so special? Put simply, the ability of qubits to be placed in superposition allows them to store previously unthinkable amounts of information. Two qubits alongside each other can be placed in four possible states (00,01,10,11), but the 53 cubits that Google used in its prototype machine provided upwards of 9 quadrillion possible states (2^53). Adding more qubits exponentially increases the processing power of the machine, allowing it to calculate in minutes what for a MacBook would take millenia.

Google was actually the first company to stake a claim to quantum advantage: In October 2019, the tech giant became the first company to claim that it had built a machine that can solve complex problems that are beyond the reach of standard supercomputers. Using a 53-qubit superconducting circuit, the team performed a calculation in 200 seconds that it claimed would take a classical computer 10k years.

Notice that we wrote “quantum advantage,” rather than “supremacy”: Quantum advantage refers to a machine that can merely do calculations faster whereas supremacy suggests that it can perform complex tasks that are beyond the reach of a classical supercomputer.

Jiuzhang has just one-upped the Silicon Valley behemoth, reaching near genuine supremacy: Rather than using super-cold superconductors, the team at the University of Science and Technology in Hefei used photons (particles of light) to solve a problem known as boson sampling, a problem that its inventors contend is impossible for a standard computer. Jiuzhang managed it in just 200 seconds.

Despite their power, it’s going to be a long time before quantum machines become useful, but development is accelerating. Although existing machines remain fairly primitive and are unable to be used for practical purposes, quantum startup Rigetti is looking to change this. The business, started by a former IBM researcher, has announced plans to build the first commercially-viable quantum machine, and is working with Edinburgh university and Standard Chartered to investigate its uses in finance, and quantum software start-up Phasecraft to apply it to design, energy and pharma.

And they’ll be getting bigger and better as we advance through the 2020s: IBM has committed to building a 1000-qubit quantum computer by 2023 and a mn-qubit machine by the end of the decade.

WANT TO LEARN MORE-

• Quantum mechanics fit for a six year-old: Wired explains quantum in five levels of difficulty. (watch, runtime: 19:27)
• Why is building a quantum computer so hard? Controlling the qubits. (ScienceMag)
• How will we use quantum machines? (McKinsey)