Ryan Sweke walks briskly through the labyrinthine physics building at Freie Universität Berlin. He knows his way around. Before joining the tech giant IBM in California, he was a postdoctoral researcher in quantum physicist Jens Eisert’s research group at Freie Universität. Now Sweke is back in Berlin. At the beginning of 2025, he was appointed to the German Research Chair at the African Institute for Mathematical Sciences (AIMS) in his hometown of Cape Town, South Africa. The Alexander von Humboldt Foundation funds several similar positions at AIMS centers in African countries. The holders of these positions spend several weeks every year at their German partner universities. For Ryan Sweke that is Freie Universität Berlin.

“This position is exactly what I’ve always wanted,” says Sweke. The AIMS institute had already inspired him as a student, when he first encountered quantum computing at a workshop there. He notes that the opportunity to collaborate with his former research group is another positive feature, and adds, “Berlin in the summer is fantastic.”

Walking through the building, he ignores the laboratories with their complicated experimental setups and laser warning signs. His workplace: a room with large chalkboards densely covered in formulas and a laptop. Sweke, who holds a doctorate in physics, designs algorithms for quantum computers. Worldwide, they are considered a key technology of the future, even though quantum computer hardware does not currently function at the level it should in order to perform smoothly.

Quantum computers are based on the laws of quantum mechanics, which describe how the smallest particles, such as atoms and electrons, behave and interact with each other. While objects in classical physics only ever assume one specific state, the tiny quantum objects can be in multiple states simultaneously and can be connected to each other over large distances, a phenomenon known as entanglement.

Limitations of Present-Day Devices

The idea of using individual atoms or ions as computing units was conceived by two scientists independently of one another in the 1980s: the Russian-German mathematician Yuri Manin, who was later director of the Max Planck Institute for Mathematics in Bonn, and the US physicist and Nobel laureate Richard Feynman. They both recognized that quantum systems could be better simulated with hardware that is itself based on quantum phenomena. Until recently, these types of computers were considered futuristic. Now machines with hundreds of computing units exist, but they are still quite sensitive. The slightest disturbances from their environment cause them to lose their quantum advantages, the very properties that make them so computationally powerful.

“Fully fault-tolerant devices don’t exist yet,” says Sweke. Therefore, research is being conducted in two directions. On the one hand, research is being done on what fault-tolerant devices could achieve in the future, and on the other hand, on what is possible with today’s small, noisy, and imperfect machines.

“A large part of my work focuses on rigorously proving mathematically where the limits of today’s devices lie,” Sweke explains. In doing so, he shatters many illusions. For a given task, one can either develop faster algorithms or prove that no faster ones exist. Sweke is particularly interested in the latter: proving which results are impossible.

It is unclear when a high-performance quantum computer might become available. Private companies and public research institutions worldwide are vying for the best hardware. Superconducting qubits, ion traps, photonic systems – each approach has its own strengths, whether in scalability or precision. Sweke’s algorithms remain unaffected: they work on all platforms. He explains, “Others translate the commands into instructions that the machine understands.” Sweke also does research on how machine learning can be improved by quantum computers.

New Materials

Can quantum computers also help us live sustainably or even stop climate change? For example, by simply calculating how and where we can reduce emissions without placing a heavy burden on the economy and households?

Sweke explains, “In mathematics, we refer to such optimization problems as ‘unstructured NP-hard.’ ” In principle, they can be solved with both classical and quantum algorithms, but very, very slowly. For these kinds of tasks, quantum computers offer no significant speed advantage.

For materials research, however, there is hope. The development of high-temperature superconductors, for example, focuses precisely on understanding and utilizing quantum effects. Here, the potential is enormous: In ordinary cables, electrical resistance generates heat, resulting in the loss of large amounts of energy. Superconductors, on the other hand, conduct electricity without resistance, but so far only at very low temperatures. If a material were found that could conduct electricity at room temperature, it could be transmitted without loss, which would be a tremendous gain in efficiency.

There are also interesting approaches for CO₂ capture. For this purpose researchers are looking for materials that can absorb carbon dioxide from the air like a sponge. This requires an extremely large surface area and, at the same time, a strong binding to CO₂ – a difficult combination. Quantum computers could help to identify materials like this more quickly.

Sweke thinks that in practice, working with quantum and classical computers will likely go hand in hand. Each computer will solve the tasks it excels at. For example, a classical computer might suggest molecules for new medical drugs, while the quantum computer checks whether a drug possesses the desired property.

Access for Everyone

Whether the new technology itself will ever be able to operate using energy efficiently remains uncertain. Sweke hopes the industry will learn from the artificial intelligence (AI) debacle. In the US, where the enormous energy demands of AI data centers triggered a nuclear power boom, reactors that had been shut down are being restarted, and new, smaller nuclear reactors are being built specifically for data centers. Sweke hopes that a combination of a critical public and government regulations will prevent similar errors with quantum computers.

Sweke stresses that, of course, quantum computers are a “dual-use” technology, i.e., usable for both good and harmful purposes, and they could also be used to develop new weapons. He sees inequality as posing a particular risk of misuse. If only one specific group has access to such technology, it is more likely to become dangerous. If access is open to everyone, the chances of improving the lives of many people with it increase.

That is one of the reasons why Sweke chose the African Institute for Mathematical Sciences as his employer. Next year, a group of young mathematicians from South Africa will accompany him to Berlin and participate in the research.

While airplane flights do generate emissions, European researchers can engage in career-enhancing exchanges without long-haul flights, whereas their colleagues from the Global South are dependent on long journeys. Sweke notes that this is unavoidable if opportunities are to be distributed fairly.

---

This article originally appeared in German in the Tagesspiegel newspaper supplement published by Freie Universität Berlin on October 11, 2025.