Quantum computing Practicality vs. Utopia: Why we need a realistic view of quantum computing

Quantum computers rightly spark imagination because their potential capabilities are enormous. While the technology is still in its early stages, expectations are extremely high. Better basic knowledge could help set realistic demands.

When IBM introduced the first commercial circuit-based quantum computer, the IBM Q System One, in early 2019, the reactions were understandably enthusiastic. However, this enthusiasm quickly turned into disillusionment among laypeople - after all, the quantum computer, equipped with 20 qubits, would not soon replace the home PC. This also lies at the heart of a false expectation: Quantum computers are very unlikely to replace conventional computers in the foreseeable future. And this is not just due to the price or their complex architecture, but the very specific tasks for which they are currently being developed. Therefore, a realistic assessment of how they actually work and what potential they have is necessary. This ultimately has a positive effect on the development of the technology. Because no matter whether private individual or company leader: The better a meaningful classification succeeds, the more targeted real added value can be created with quantum computing.

Eliminate false assumptions

The assumptions of most people about quantum computing are still shaped by their own everyday experiences with conventional computers. Here, performance is often equated solely with speed, i.e.: how many arithmetic operations can be solved in a certain period of time. However, quantum computing relies on a number of different computing models based on quantum physics laws. A quantum computer operates with qubits, which can consist of ions or photons, for example, and which are subject to various quantum phenomena. These include superposition, in which qubits can assume arbitrary superpositions (linear combinations) of states 0 and 1, and that they can be entangled with each other (Entanglement). As a result, the state change of one qubit has an instantaneous, non-causal effect on the qubits entangled with it.

Neuland betreten

Sichere optische Datenübertragung mit Quantentechnik und Licht

Entanglement leads to an effect that is essential for quantum computing and differs significantly from the way conventional computers work: a quantum system can generally only be understood as a whole and not in its individual parts. This is vividly illustrated by a thought experiment by theoretical physicist John Preskill: In a book based on quantum laws, there would be information on the individual pages - but we could not read it out. This is because most of the information is stored in the correlation, i.e., the entanglement, of the individual pages with each other, which is why such a book would only provide usable information as a whole. Having to always consider the whole may sound inefficient at first. However, it has a decisive advantage for problem solving: after all, to put it simply, it is not an intermediate solution but the final result that is often of greatest interest.

Do all roads lead to Rome?

Why looking at the whole picture, as well as various other phenomena such as superposition or interference, ensures that quantum computers can solve certain problems faster, more optimally or more energy-efficiently, is shown by another example: If there is an apple in one of four drawers in a chest of drawers and it needs to be found, the Grover algorithm can, put simply, obtain information about the contents of all drawers simultaneously. On the other hand, a conventional computer would have to open up to three drawers to definitively determine where the apple is located. The example represents an unstructured database search with four entries. The interaction of various quantum phenomena such as superposition, entanglement, and interference ensures that with the help of a quantum computer, only the time needed for querying a single database entry is required.

There's an old saying that all roads lead to Rome. But what if you're looking for the ideal route for a trip to Italy that also includes stops in Naples, Florence, Bologna, and a number of other cities - with each location being visited only once, the journey being as short as possible, and ending up back at home? This is a well-known combinatorial optimization problem - or the "travelling salesman problem". A quantum computer could potentially compare all theoretically possible routes simultaneously to ultimately find the optimal route. Comparable complex challenges are also found in the design of microchips or the planning of global logistics chains. Traditional computers can and are used to solve these problems, but the higher the number of constraints, for example, the greater the effort - and in some cases a solution is even impossible for them. Quantum computers, on the other hand, have the potential to handle this level of complexity better.

Use it where it makes sense.

The performance of conventional computers continues to increase at certain intervals. Thanks to some promising approaches, such as three-dimensional integrated circuits that are supposed to replace classical semiconductor technology, an end is not expected anytime soon. At the same time, they will continue to work with 0 and 1 - the only acceptable states of conventional bits. This is perfectly sufficient for numerous tasks. Therefore, it is assumed that they won't be replaced by quantum computers in the foreseeable future, but rather that these will be a systematic addition. After all, it's important to choose wisely, depending on the task at hand, whether a classic or a quantum computer is better. For many tasks, which are almost ubiquitous in today's world, classic computers will probably remain the first choice for a long time.

Consequently, it is very unlikely that quantum computers will take over tasks in the future for which conventional computers are perfectly adequate. Rather, the focus will be on pursuing a hybrid approach - i.e., the interplay of classical and quantum computers. At the start of a project, the first step would be to identify requirements for which quantum computing is relevant. An inherently promising candidate is the aforementioned field of combinatorial optimization. While not all individual criteria have been completely clarified yet, the following are likely to play a role: the complexity of a problem, for example, based on the number of side conditions, and bottlenecks in the field of machine/deep learning, for example in terms of linear algebra or the expressivity of a model. The latter roughly describes how complex a task can be that a model is capable of learning.

Accompany to understand and benefit.

The insights of quantum physics have long since arrived in practice. In fact, they have been used for decades in a range of technologies, including in electronics, chemistry, and digital technology. In order for quantum computing to fully realize its potential and be used effectively, it needs companions like Vinci Energies, who are both users and integrators. They will help in the future to promote a realistic understanding of the technology, identify suitable fields of application and finally carry out the practical application. This way, it is possible to realistically and purposefully classify the potential of a complex and exciting field like quantum computing and solve exactly those problems for which the technology represents an optimal solution.

Straight to the goal.

The potential of quantum computers is to solve highly complex tasks faster, more optimally or more energy-efficiently, where the goal is an end result without intermediate steps. However, there will continue to be many other problems for which classical computers will be completely sufficient to solve. It is very likely to expect that quantum computers will tackle tasks, the solutions of which classical computers will then continue to work with. Such a hybrid model would most sensibly exploit the potentials of the two, simply put, computational models in order to optimally use the respective strengths and thus resources. We should abandon the idea that it would be sensible to replace conventional computers with quantum computers in the foreseeable future.

This article originally appeared on our partner portal Industry of Things.