What could quantum computing do for retailers?

In the second of a three-part series examining the rise of quantum computing, dunnhumby Data Scientists Ross Williams and David Hoyle discuss some of the practical applications of this fast-moving technology – as well as the current limitations on its potential.

As mentioned in our previous post in this series, one of the key challenges for any organisation when it comes to quantum computing (QC) is the need to understand its commercial potential. What can it actually help us do?

To answer this question we need to understand more about the technology. Moreover, starting on the quantum learning curve now will help a business demystify this complex landscape and separate the hype from the genuine opportunity.

Quantum computers use a basic unit of information known as a qubit. Qubits are the quantum equivalent of the bits found in the computers we use in everyday life. And while bits must have a defined value of either 1 or 0, qubits can exist in any combination of these two states. This is what is known as a “superposition”. This superposition flows through any operations that a quantum computer performs, meaning that just a single 64-qubit quantum register can encapsulate the results of 2⁶⁴ ≅ 1.84 ×10¹⁹ calculations.

While we do not go into the details here* of why QC is not massively parallel computation, this simple example hints at the capabilities offered by quantum computers. And that promise is the ability to perform certain mathematical calculations dramatically faster than currently possible. Calculations so hard that they could never be solved through traditional means.

We’re getting increasingly close to realising that promise. In 2019, Google’s 54-qubit Sycamore processor performed a 200 second computation that would have taken the world’s fastest supercomputer (at that time) around 10,000 years to achieve¹. While IBM contested² Google’s 10,000 year claim, the idea of “quantum supremacy” – the point at which a quantum machine can perform tasks that would fundamentally elude a traditional one – is tantalising nonetheless.

That concept should be of particular interest to anyone who deals with data science at scale and, unsurprisingly, it is.

One class of mathematical problems familiar to retailers is that of optimisation. These come up everywhere in grocery retail, including routing delivery vehicles, choosing promotional strategies, store arrangement and the allocation of shelf space. Today’s solutions are often limited by computational power, and consequentially limited in size and scope. These optimisation tasks also happen to be the sort of challenges for which today’s quantum computers could prove to be game-changers. So why “could”, then, and not “is”?

Part of the reason is that we’re still at the very early stages of commercial QC. Quantum machines remain relatively scarce, and those that can be leased on an as-a-service basis come with limitations. The limitations come in two key forms: the finite number of qubits, and noise which causes a build-up of errors in a quantum calculation. Together they restrict the processing power available.

Beyond those technical issues, industry as a whole is still in the process of working out the exact language of quantum computing. This isn’t just in relation to the coding languages that we use, though that is an undoubted issue, but also in terms of our ability to frame problems in a way that allows QC to tackle them effectively. Essentially, we need to get better at mapping business problems onto quantum machines, as well as learning which of those problems will actually benefit from all of that computational power.

This, of course, is only the state of play today. If the last decade has taught us anything, it’s that technology tends to move very fast – particularly when the likes of Amazon, Microsoft, and Google come on board. We might not be there just yet, but there is little doubt that we are moving into an era in which businesses of all kinds will be able to create genuine competitive advantage from QC.

In the third and final post in this series, we’ll be taking a detailed look at the factors that need to evolve in order to propel us into that new age.

*This blog discusses themes covered in greater depth in our new report on quantum computing: “Demystifying quantum computing for commercial companies.” Download a complimentary copy of the report here.

¹ Quantum Supremacy Using a Programmable Superconducting Processor – Google AI Blog, 23^rd October 2019
² On “Quantum Supremacy” – IBM Research Blog, 21^st October 2019