21 July 2024

The holy grail of algorithms

Mari Carmen Bañuls is a physicist and a computer scientist. Her scientific career began in particle physics. At the Max Planck Institute of Quantum Optics, she now develops algorithms for quantum simulations.

Mari Carmen Bañuls is a physicist and a computer scientist. Her career started in Valencia, Spain, with a fascination for particle physics, leading her to computer science and finally to the Max Planck Institute of Quantum Optics. Here, she is now a group leader in the “Theory” department and works on the development of algorithms for simulating complex quantum systems. Driven by curiosity and openness, Bañuls emphasizes the importance of teamwork and exchange within the scientific community. To encourage this, she is involved both as a mentor for doctoral students and for the institute as a whole.

Mari Carmen Bañuls is a modest person. It is almost impossible to elicit any immodesty from her in conversation. For example, take the case of the "Orden del Merito Civil". In 2015, King Juan Felipe of Spain awarded this medal to 38 Spanish citizens in recognition of their services to the Spanish nation. One of those honoured was physicist and computer scientist Bañuls.

If you ask her why she was honoured, she comes up with a couple of reasons: The king wanted to demonstrate proximity to citizens on the occasion of his first anniversary on the throne. That is why he not only honoured well-known personalities, but also completely normal, unknown people. He wanted to honour someone whose work has taken her or him abroad. When Bañuls is told that her excellent research, which is respected all over the world, has certainly also played a role, she laughs – without confirming.

Mari Carmen Bañuls has been researching at the Max Planck Institute of Quantum Optics in Garching for 19 years. Here she is a group leader in Ignacio Cirac's "Theory" department. Bañuls discovered her love of physics at school and therefore decided to study this discipline in her home town of Valencia. Particle physics was her initial focus: "The professors' lectures on this subject simply inspired me," she says, "and as a young person, you often do what your teachers inspire you to do."

From particle physics to quantum physics

Then as now, she was particularly fascinated by one phenomenon: Quantum entanglement, i.e. the direct effect of quanta on each other, even when they are spatially separated. "This relationship between the parts of a quantum system is completely counterintuitive," says Bañuls, "but it has dramatically real effects on the results of local experiments." This phenomenon is one of the reasons why quantum systems are incredibly complex; so complex that they can only be understood in part using existing methods. Scientists are therefore feverishly researching new algorithms with which quantum systems can be simulated reasonably realistically. Spoiler alert: Mari Carmen Bañuls is working on developing such algorithms at the Max Planck Institute in Garching. But it has been a 30-year journey until then. During this time, she has often been able to demonstrate her curiosity and her ability to familiarise herself with completely new specialist areas at the highest level.

To understand this, let's first go back to Bañul's place of study, Valencia. In 2000, particle physics did not seem to be the ideal field for her further scientific development: too mainstream, too few vacancies. Together with like-minded people, she found herself in a small study group that discussed exciting topics in physics and related fields: "I wanted to be closer to the application," says Bañuls: "And since I've always been interested in computers and programming, I got into computer science." This sounds almost like a hobby now, and Bañuls – willingly – sacrificed a lot of free time for it.

Tensor networks: simulating complex systems in small segments

With both – particle physics and computer science – the Spaniard had all the knowledge and techniques she needed to embark on the search for her personal holy grail: an algorithm with which complex quantum systems can be simulated better than before. The tool that she is constantly developing and optimising for this purpose are so-called tensor networks. Tensor networks are a mathematical method by which certain aspects or states of complex systems can be picked out from the totality of all possible states in order to better describe, simulate and understand them as examples for the entire system.

To the layperson, it may sound surprising that Bañuls sees tensor networks as something closer to application than, for instance, particle physics. But the 50-year-old is good at explaining it: She cites high-temperature superconductors as an example. Normal superconductors conduct electricity without loss at temperatures close to absolute zero (minus 273°C). High-temperature superconductors "only" need around minus 140°C to demonstrate superconductivity. "With superconductors, thanks to theoretical models and conventional simulations, we understand quite well what happens inside them and how the electrons behave," says Bañuls: "With high-temperature superconductors, the conditions are too complex. We need new methods for this."

Tracing the secrets of quantum systems

Tensor networks make it possible to observe precisely defined states of electron systems and to answer specific questions relating to these states. This brings physicists a step closer to the secrets of high-temperature superconductivity. The same applies to quantum systems, such as those on which quantum computers are based. With tensor networks, researchers like Bañuls can "look at individual families of states of the qubits", as she says. The group leader from the Theory department is particularly interested in the temporal dynamics in which these states evolve – and thus ultimately determine the computing power and computing results of quantum computers.

For Mari Carmen Bañuls, application-orientation does not mean that she investigates a physical phenomenon in order to make it usable for a specific technical solution. Her work is and remains theoretical. However, she systematically develops theories such as those on tensor networks in order to address issues that need to be resolved for the technical application of complex systems. Her colleagues at the Max Planck Institute and in the large scientific community outside the Institute appreciate this. "I work with many groups that are not making progress with their standard methods," she says. Tensor networks are then often a new approach with which the problems can be solved in a different way after all.

Commitment to the scientific community and to the team

Listening to Bañuls, you get the impression that she sees herself more as a member of a large scientific group than as a group leader. So she was also the Ombud Person at the institute for many years, a confidential go-to contact in questions relating to scientific misconduct. Bañuls explains her commitment with the belief that "you also have to give something back to the institute”.

Despite her commitment to the community, her direct team is also very important to her: "Students who want to do a doctorate in the "Theory" department will find a good atmosphere here. For me, it's important that they are open to new challenges and enjoy interacting with other scientists. Then the work here is a lot of fun." Bañul's group usually consists of four to five doctoral students. After their scientific training, they are also valued experts in complex systems; such as in start-ups like the spin-off from the Max Planck Institute of Quantum Optics, "planqc", which is working on the development of quantum computers: "Colleagues like to hire our former doctoral students," says Bañuls happily – a clear proof that the Spaniard is pursuing very abstract theory for very practical application.

Source: MPQ Website