Quantum computers have garnered a reputation that borders on the mythological. The Internet is filled with fantastic stories of what these machines will someday be capable of. Indeed, large companies such as Google and states such as China and the USA are investing enormous sums of money in developing quantum information technologies. Germany is also getting involved; the federal government recently invested two billion euros in furthering quantum technology research, while the federal state of Bavaria has invested 300 million euros in the development of Munich Quantum Valley. But what will quantum computers really be capable of once the technology has been perfected? What will remain impossible? What is realistic and what is mere hype? These are just some of the questions that Professor Jens Eisert, researcher at the Dahlem Center for Complex Quantum Systems, Freie Universität Berlin, deals with in his work.

Not Your Conventional Computer

“Quantum computers are still very much in their infancy,” says Professor Eisert. “First and foremost, they are computers. That means they can solve tasks just like a standard computer. But the difference is in how they function – and there are certainly some differences!” Conventional computers deal with the tasks they are given by using yes/no decisions as basic logical expressions – those famous ones and zeros that make up “bits.” The quantum bits used by a quantum computer – qubits for short – also use a two-state system that you could refer to as “one” and “zero.” However, this is where the laws of quantum mechanics come into play. “This is, after all, the theory that most precisely describes the physical properties of nature – on the scale of elementary particles like atoms and photons,” says Eisert. Since we experience life on a larger scale than this, our reasoning and perceptions are not necessarily used to considering the characteristics of the quantum world. As a result, the particularly sensitive quantum effects that are susceptible to the ever-present interferences of the macro world may seem particularly strange to us. In this regard, the phenomenon of entanglement might come across as especially odd. Here, a group of particles can form a single unit in terms of their quantum state, even when separated by a large distance. Despite this distance, any change made to one of the particles is immediately “noticed” by the other members of the group as far as correlations are concerned.

Quantum computers make use of principles such as these. For example, a qubit can be in a zero or one state, or in a superposition of the zero and one states at the same time. This allows information to be encoded in a strikingly different way. Additionally, the states of multiple qubits can be superposed and entangled with one another. This would mean that huge amounts of computational power could be generated with just a few qubits. For this reason, quantum computers should, in principle, be able to solve certain types of problems that would overtask even the most modern supercomputers. These are tasks that would require the processing times and memory of classic computers to expand exponentially – like the explosive spread of a pandemic, but for computing tasks. Some of these problems are technologically very interesting. It is hoped that quantum computers would be able to elegantly solve many of these with just a few qubits and within a feasible amount of time. “This has actually already been proven to be the case for certain paradigmatic problems,” says Eisert.

The Encryption Dilemma

To illustrate this, Professor Eisert uses the example of multiplying numbers. “Multiplication is a simple task. Even my seven-year-old daughter can work out that 3 x 5 = 15,” he says. “And if I have to multiply big numbers, I can do that with the technique that I learned at school – using pen and paper.” However, if you already have the answer and want to find out which prime numbers multiply together to make it, things start getting complicated. Also known as “prime factorization,” this process can quickly push even a modern supercomputer to its computational limits. For exactly that reason, this process currently forms the basis of data encryption. It is secure because even the best computers cannot crack this kind of code quickly enough to have a significant impact on whatever is being kept confidential.

Quantum computers, however, would be able to carry out this calculation with just a few qubits, thanks to their exponential computing power. This was already theoretically proven by American computer scientist Peter Shor in 1994. Shor’s algorithm was one of the first quantum computer algorithms to demonstrate the superiority of quantum computers over standard computing. In principle, programming a quantum computer is not much different to programming a conventional computer. “Here, complex functions are also broken down into basic ones,” explains Eisert. “Conventional computers use logical decisions, or logic gates. Accordingly, quantum computers use quantum gates.”

The difference is in how these “decisions” are made. While conventional computers carry out calculations one after the other, quantum computers are able to carry out calculations all at once as soon as the quantum gates are programmed thanks to their ability to superpose qubits. This ability is also referred to as “quantum parallelism.” However, it is important that the entangled qubits remain perfectly isolated and undisturbed while carrying out their work. Taking a peek would be a big no-no; this would immediately stop the computing process. You can imagine the process a bit like cooking with a pressure cooker. You put the ingredients in and only open it once your dish is – hopefully – done.

However, a pressure cooker usually has some kind of display providing information on the pressure levels and cooking process, which would be unthinkable for a quantum computer. After all, in quantum mechanics, measuring something alters the thing itself. “Imagine measuring the length of a table and, through being measured, the table becomes longer or shorter,” says Eisert. In our macroscopic world, we do not have any absurd experiences like this – the amount of interference produced by our environment blocks out these quantum effects.

Handle with Care

This sensitivity to even the tiniest disturbances is the Achille’s heel of all quantum computers. The field of physics has developed fascinating techniques that allow us to capture single atoms with laser beams and assemble them into qubits. Electrically charged atoms – or ions – can be kept isolated for weeks in electromagnetic traps. This is a technique that allows for simple quantum computing operations to be carried out. Using tiny, superconducting electronic circuits to connect qubits is another way to do quantum computing; this is a technique currently being explored by companies such as Google and IBM. However, no one has been able to isolate qubits from environmental interference to the extent needed to produce quantum computers with hundreds, thousands, or even tens of thousands of qubits. It will be a long time until we see the first fault-tolerant quantum computers – no one can predict when or even if they will become a reality.

Professor Eisert says, however, that the first “dirty” quantum computers are already here. With “dirty,” he is referring to interference from noise in the environment that reduces the lifetime of quantum information and makes quantum calculations increasingly “muddy” over time. “Despite their imperfections, these first quantum computers are invaluable,” he explains. They allow researchers to gauge what might be possible in the future. Currently, a lot of work is being done to develop specially adapted quantum algorithms that will serve as “software.” With regard to hardware, researchers like Eisert are pursuing the idea of dividing tasks into one part that can be processed by a conventional computer, and one part that requires a quantum computer. This would result in a “hybrid” computer architecture. According to Eisert, “This would mean using a larger, conventional computer as an environment into which we can insert our little quantum computer.”

Staying on Track with Quantum Computers

When reflecting on real-world applications that could benefit from quantum computers, Eisert presents the Deutsche Bahn, Germany’s national railway company, as one example. The Deutsche Bahn sometimes struggles with scheduling problems and would like to resolve this issue using new information technology. “Generating timetables is a highly complex optimization problem,” says Eisert. Perhaps the most well-known situation is that of the commercial traveler who wishes to visit multiple places and is seeking the quickest route between each one. The more places they want to visit, the more complicated the request and the more overloaded the computation. This is an area where quantum computers could be of assistance, which is why Jens Eisert’s team is in contact with the Deutsche Bahn.

Clearly passionate about his work, when describing his field of research, Professor Eisert says, “It’s like a thriller.” Nobody knows what breakthrough is coming next.

This article originally appeared in German on September 27, 2022, in the Tagesspiegel newspaper supplement published by Freie Universität Berlin.