Quantum computers are one of the key technologies of the 21st century. They are intended to revolutionize the computational performance of computers. Researchers also hope to gain new insights in science, for example in high-energy physics or quantum chemistry. But what exactly will quantum computers be used for and what might they look like? This is the focus of the interdisciplinary research team of the first Einstein Research Unit (ERU) "Perspectives of a quantum digital transformation: Near-term quantum computational devices and quantum processors" of the Berlin University Alliance. Professor Jens Eisert is one of the main applicants. He conducts research as a physicist and mathematician at Freie Universität Berlin.

Mr. Eisert, for many people quantum computing is an abstract keyword. Can you briefly explain what exactly this is about?

First: Quantum computers are computers. They solve the same kind of mathematical problems as conventional computers. But quantum computers can solve certain issues far more quickly. Certain mathematical problems are so complex that it is completely unrealistic to solve them in a tolerable time on conventional computers.

An example: Two numbers, even large numbers, can be multiplied with paper and pencil. But if I take the product of two large numbers and want to figure out which numbers were multiplied, it gets tricky. The best-known algorithm takes some time to do this, which grows exponentially with the length of the input. It is practically impossible to solve this task. However, a quantum computer can solve it quickly. That is what makes it so exciting. Unlike classical computers, it handles information based on quantum mechanical laws. This means that memory contents on these computers can, virtually, contain multiple, superimposed values, on which computing instructions have a simultaneous effect. Therefore, for certain problems, the quantum computer is a supercomputer that solves tasks not only faster, but in a novel way, through a fundamentally different kind of computation that shakes the foundations of what computation means.

How far advanced is the development of quantum computers? What does the German research landscape look like?

Ready-to-use quantum computers are not yet available. The idea of these computers is not new, it dates to the 1980s. It existed primarily as an extremely inspiring thought model. This has offered a new angle to look at quantum physics, for example. Novel simulation algorithms, so-called tensor network methods in solid-state physics, are inspired by ideas from quantum computing. Only more recently have attempts been made to build such devices. This development was strongly promoted by the quantum industry and US-American companies like Google, IBM, Rigetti, and PsiQuantum. The development is quite apace and fascinating.

As part of our research group at Freie Universität Berlin, we are working with Google AI and have characterized the quality of their chip: So, we have understood how exactly the various parts of this chip work. And this chip is not a small one: It already contains 53 qubits – which is not bad. Qubits are the smallest computational and memory units of quantum computers. Recently, there have only been chips with 3 to 5 qubits; the next generation will apparently have thousands. With this, one can then already make applications that could be practically relevant according to all expectations. In research, the hope is that such system sizes are already sufficient to explore interesting applications beyond the capabilities of classical computers. Exactly what these are is also one of the tasks that ERU Quantum will address.

What specific questions do you want to address with ERU Quantum?

Quantum computing is all the rage, but we ask: for what can quantum computers really be used? We want to look closely, unreservedly, and without hype, at which tasks are suitable for this. People are flooding us with problems like routing flight schedules or car manufacturing schedules and they want to do it faster on the quantum computer. However, this can only solve specific, highly structured tasks. So which ones are realistically better to solve than with classical computers?

In this context, we also ask an experimental question: how to build parts of quantum computers with quantum optical architectures, i.e., using light? Such experimental platforms are receiving significant attention in Berlin, and so it is only logical to pursue such approaches.

Quantum computing attracts a great deal of interest, but also false expectations. We take an extremely interdisciplinary approach. Berlin has a great deal to offer in this area: We bring together experts from the fields of computer science, signal processing and machine learning. That is a unique feature.

You have raised the BUA's first Einstein Research Unit. Why is the format particularly suitable for your research project?

The exploratory format of ERU allows us to ask new questions in a fresh way and fits well with our project, as does the focus on interdisciplinarity. However, the time frame could be significantly longer. Research should be sustainable, and we aim to continue the project beyond the two years.

What is exciting about this format for you personally?

I find interdisciplinarity exciting. Machine learning is a trend, quantum computing now too. This is what cries out for the two disciplines to be combined. Is there a quantum advantage to this? This can mean that the problem is calculated faster or that fewer data is needed for the calculation. We have demonstrated that, in fact, quantum machine learning does exist, and there are cases where the quantum computer performs computations exponentially faster than a classical computer. That is cool!

Unfortunately, the whole thing is not yet as practical as we researchers would hope, but it is important to have made a start. This was only possible because we knocked heads together with the Einstein professor and computer scientist at Technische Universität Berlin, Jean-Pierre Seifert, who is also an ERU Quantum main applicant, and thought about how to combine the two disciplines. That is also the remarkable thing about Berlin: there is a specialist for every unique subject. None of us could have done it alone. That is what makes ERU different: We have expertise from unusual perspectives and can make fresh contributions.

Questions asked by Ina Friebe