Matthias Troyer on his pivot to Microsoft — and the importance of mid-career recognition

In 2017, computational physicist Matthias Troyer left a successful career in academia for a position leading Microsoft’s quantum computing division. “I wanted to build something,” he says — like the company’s first quantum computer, and the software to run it.

More recently, Troyer — elected to APS Fellowship in 2011 — decided it was time to help APS build something new, too. Through his own generous donation, a matching donation from Microsoft, and donations he helped secure from other APS members, Troyer helped establish the Charlotte Froese Fischer Mid-Career Award in Computational Physics.

The award, which just closed its inaugural application round, will annually offer one recipient a $5,000 stipend and a travel grant to deliver an invited lecture at an APS meeting. It honors the late Charlotte Froese Fischer, a Canadian-American applied mathematician, computer scientist, and physicist who published more than 300 articles on atomic theory, and whose prediction of the existence of the calcium anion led to its experimental discovery.

APS offers a range of prizes, including many for students, early-career scientists, and established scientists. But Troyer most liked the idea of supporting a new mid-career award. “A prize can actually have a bigger impact when it can still shape a person’s career,” he says.

APS News spoke with Troyer to learn more about his work, his involvement in APS, and his inspiration to support its new mid-career award. This interview has been edited for length and clarity.

What core questions have driven your work as a computational physicist?

As a student, I was fascinated by trying to understand how nature works, from the smallest sub-atomic particle to the universe. When I got into computing and, with other students, developed our school’s first software infrastructure, I wanted to bring these two interests together — to see how we could simulate nature with computers. Of all fields, I found quantum physics the hardest, so I started my university studies with physics. But I kept my interests in materials, chemistry, and computing.

For my master’s thesis, I was offered a topic where I could use a Cray X-MP supercomputer. It was clear that this was what I wanted to do. For years, my focus was then on using computers to solve physics problems. If you combine the best computers with new algorithms and interesting physics problems, you can make real and fast progress — finding new problems to solve with the new algorithms and new hardware.

In 1997, you moved to Japan for a postdoc, where you had access to a supercomputer. How have supercomputers changed since?

I moved for the people, the scientists, and the physics in Japan. On top of that, I had access to a machine that had a performance speed of 220 giga flop [1 billion calculations per second]. Today, the biggest supercomputer in the world is 10 million times faster. That machine in Japan — its power is like my cell phone today.

Plus, the machines have gotten even bigger. With the end of Moore’s law, people started to get more performance by scaling out to more processors, more CPU cores, rather than making the processors faster. For example, the machine that I worked with in Japan had 1,024 cores. Currently, the fastest supercomputer has 11 million cores. So it’s 10 million times faster, but it also has 10,000 times more CPU cores — these machines have really grown in size.

Pivoting from academia to Microsoft was a major mid-career transition. Say more.

As one evolves and goes through one’s career and pivots and changes, one learns one’s own strengths and passions, and then finds the right role over time. I’ve sometimes thought whether I should have done things differently. Should I have pivoted earlier, or not? But in the end, whatever I’ve done, I’ve learned new things, built up skills, and had experiences that prepared me for the next phase in my career. In hindsight, there’s a strong story to be told that one could not have predicted.

You've been an active member of APS for more than 30 years. Why?

I first got involved with APS as a graduate student back in 1994, when I went to the March Meeting. My talk was on ‘two-leg ladder models’ — models of two coupled chains of strongly correlated electrons. And it was the first talk Monday morning. The audience was mostly empty. [laughs]

Even now that I’ve moved to a company to build quantum computers, APS still plays a big role for me. The journals are a reliable and well-respected venue for publishing results. Also, the meetings are for more than talks. I go to them for networking and finding top students and postdocs to fill our open positions. I like to engage with other physicists — innovative ideas come from academia.

You’ve been the recipient of several prominent awards. Why do awards matter?

In my career, the awards I received — like the gold medal at the International Chemistry Olympiad in 1986, the ETH medal for an outstanding doctoral thesis — helped with reputation, helped with building my career. Awards help because they are a form of confirmation. Just being able to mention the awards on one’s CV helps a person’s career.

What’s the significance of APS offering a new award in computational physics?

As a computational physicist, one is always between the computational science community and the physics community. To make progress in that field, one has to be an expert in both computing and physics. The challenge is that one gets evaluated for careers only by their impact, for example, on physics. But that is a hard career, and it has been a challenge for many colleagues to get recognition in physics as a computational physicist.

Now, computing is becoming more important across all fields of science and engineering. So, to make progress in physics, we have to help get computation in physics recognized, and that’s where a prize like this can help.

Why is the invited talk an important aspect of APS’ new Charlotte Froese Fischer Award?

As a student, a postdoc, and a young faculty, the main impact of APS on my career was through networking at the meetings and invited talks. Invited talks are a sign of recognition, and they help with one’s career. As part of the award, there’s an invited talk for the recipient. Typically, one can only give one invited talk at an APS meeting. But if you have an award, it’s possible to give two, the standard invited talk plus the award talk. That’s the only way one can give two invited talks at one meeting — for twice the career impact.

What scientific questions might we be able to answer with supercomputers and AI that once would have been a dream?

Soon, we can simulate materials and chemicals very accurately, predictively. We can already do it in many cases. What’s more, the dream is that with ‘agentic AI,’ I can just talk to a computer and ask it to do certain calculations. But really, I think most of the things we do now feel like a dream!

The Charlotte Froese Fischer Award was funded in part by a corporate match from Troyer's employer, Microsoft, one of many companies with donation matching programs.