Lukas Burgholzer enjoys his work so much that he sometimes finds it difficult to step away from it. The many challenges that still need to be solved in order to develop software for quantum computers are an incentive for the young computer scientist. His team, with whom he gets along so well, is no less of a motivation.
By Maria Poxleitner
Eleven o'clock at night. Lukas unlocks the door to his apartment in Linz and puts down his sports bag. The two hours of tennis did good – unwind, clear the head. Now the young computer scientist is drawn back to his computer. Yesterday, they spent hours tearing apart the program in the work group, but failed to find the crucial error. But just now an idea struck him. He opens his laptop and begins meticulously combing through one line of code after another.
"I just can't help it," admits Lukas Burgholzer with an apologetic smile. When he has an idea for a solution to a problem, he wants to try it out right away – no matter how late it is. "It just keeps me up at night," laughs the 29-year-old. Separating work and free time is not always easy for the aspiring postdoc, but this separation may not be necessary if work is also a hobby. And it definitely is, Lukas emphasizes.
Studying mathematics, not computer science, was the first choice. During his math bachelor's degree at Johannes Kepler University in Linz, he attended his first programming courses. Before that, he had never learned how to do so, Lukas says. In the end, however, he enjoyed programming so much that he completed a bachelor's degree in computer science in parallel to his master's degree in mathematics.
One of his computer science professors, his later Ph.D. supervisor Robert Wille, inspired Lukas so much that he decided he wanted to write his bachelor's thesis with this very professor. "Optimal Mapping of Quantum Circuits to the IBM QX Architectures" was ultimately the title of the thesis. That it became a quantum-computing topic was more of a pragmatic decision; it simply "fit in well" with Robert's working group at the time, Lukas says. Until then, he himself had never come into contact with the subject of quantum computing.
In the meantime, after four years of Ph.D., he is fully immersed in the topic and books, which many physics students also pore over when they want to familiarize themselves with quantum information theory, are piled up on the desk in Lukas' Munich office. "I'm sort of researching the software for the computers of tomorrow, so that you can use them like the computers of today," the prospective postdoc explains when asked for a brief summary of his research. No one who works with Excel or plays computer games needs to understand what goes on in their computer's processor, how that processor is controlled, or how to write programs that do something useful on that processor, he says. All the numerous steps between the user and the hardware are automated – for which a lot of software has been developed over the last 50 years, Lukas adds.
Position
Postdoc
Institute
TUM – Chair for Design Automation
Q-DESSI & QACI
Degree
Mathematics & Computer Science
Lukas develops methods and tools that help automate important steps in software development for quantum computers. One of his main focuses is on verification methods, i.e., the development of automated procedures that check programs for functionality. For example, an automated "equivalence check" can eliminate a compiler as a source of errors by verifying that source and machine code have the same functionality.
Compilers are an important building block in this so-called "software stack". Applications today are written using "high-level languages" such as JavaScript, C++ or Python – programming languages that can be used to express highly complex functionalities. The programmer's code, called source code, is then translated into a language that the processor understands, and ultimately implemented with logic gates. This "translation" from programming language to machine language, known as compilation, runs completely automatically. In quantum computing, on the other hand, much of this is still done "with pen and paper," says Lukas. "We're at such an early stage that applications are largely programmed in such a way that people actually indicate sequences of individual quantum gates to be executed. Of course, that's not at all feasible or scalable in the long run," he explains. With today's computers, where there are billions of transistors on a chip, that would also be unthinkable. Developing automated solutions and ultimately ensuring that users can use quantum computers without having to be quantum or IT experts – "that's our mission for the probably longer-term future."
But it is precisely the fact that there is still so much to do on the part of computer science that drives Lukas: "I find the perspective of software development for novel systems, such as quantum computing, extremely exciting. You are at the forefront of science and can make big contributions in a relatively short time."
Initially, it was not at all clear to the young computer scientist that he wanted to go into research. After completing his master's degree in mathematics and while still writing his bachelor's thesis in computer science, he was already working for a company. But when Robert Wille offered him a position as a doctoral student, the decision was made quickly and the job was quit. When his doctoral supervisor moved to TUM during his Ph.D., it was clear to Lukas that he would go along – even if he wouldn't turn his back on his Austrian hometown to do so. Lukas still sees Linz as the center of his life: his girlfriend, a large group of friends from his childhood, the city where he feels at home. He will therefore often commute between Munich and Linz.
Lukas sees the fact that his research group is now based in Munich as a great opportunity: "In Linz, we were one of the few groups working on quantum computing. In Munich, there are lots of people in the vicinity who are doing research on quantum computing and with whom you can exchange ideas." But this exchange also demands a lot from the scientists. "When we as computer scientists talk to physicists, it takes a lot of time to make sure that both parties mean the same thing," says the 29-year-old, thinking back to meetings with several MQV scientists that focused on compilation. Everyone speaks a different language, but if you don't talk to each other, you run the risk of solving the wrong problems, Lukas warns. Because a compiler acts as a translator from programming language to machine language, it makes a difference whether the processor uses superconducting qubits or qubits encoded in atoms. For example, when compiling programs, one has to be careful that, depending on the hardware technology, only certain quantum gates are supported or that not all qubits can "talk" to each other, the young computer scientist explains. "In order to write a compiler that compiles to a particular hardware in the best possible way, we need characteristics of the physical systems that we can put into our models." You have to try to understand the physics a little more, he says, while the physicists have to create levels of abstraction and not throw around physical details. "We have to somehow come together on this, because isolated solutions will not lead to success."
At noon, Lukas meets with Nils, Aaron and other doctoral students from the group. With his hands in the pockets of his shorts, he walks with his teammates from the TUM campus on Arcisstrasse over to the cafeteria. Since a bet in his school days, he always wears shorts, even in winter, unless there is a formal occasion that prevents him from doing so. It's almost like a trademark, Lukas says with a laugh. Perhaps also a sign of his optimistic streak. The sun will certainly shine warmly enough on this otherwise still chilly spring day.
During lunch break, there is finally time for a personal exchange. Otherwise, the days in Munich are filled with numerous meetings and discussions. The fact that the small group from Linz, consisting of Robert Wille and three doctoral students, managed to move to Munich in the last one and a half years and build up a group with now more than 15 doctoral students is something he is particularly proud of, says the young computer scientist. "All of them are mentored and deliver scientific contributions that have impact. I think I've definitely played my part in that." He has always benefited from the regular and direct exchange with his doctoral supervisor over the past few years, he adds. Now he is in the role of the supervisor himself and wants to pass that on to his doctoral students as well: to be available at all times for questions and concerns, to show that someone is standing behind them, but also to be critical and try to tickle the best out of people – and also out of oneself. "I can get on the other people's nerves sometimes," Lukas says, laughing mischievously.
After lunch, the group sits down for coffee in the department kitchen and shares stories about past and upcoming conferences. Singing karaoke in Tokyo or mastering rocky climbing passages to scale Bear Peak in Colorado. These shared experiences that accompany many a conference are incredible opportunities, Lukas enthuses: "It unites the team, and I'm really grateful that this is possible through work." He also sees it as appreciation for the hard work they do together. Finally, it is also he who ends the break and calls to continue working. After all, papers have to be written before you can go to a conference, he notes.
He has never really made any big plans for the future, says Lukas, but rather lived from moment to moment. "Everything has always worked out in some way and then fit well". In the meantime, however, he sees his place in research. "I would be very happy to continue working in this group. Actually, I wish everything to continue as it is right now."
Published 21 July 2023; Interview 17 March 2023