Interview with Margaret Martonosi

Originally Printed in the Summer/Fall 2013 Newsletter

Margaret Martonosi is the Hugh Trumbull Adams ‘35 Professor of Computer Science at Princeton University, where she has been on the faculty since 1994. She also holds an affiliated faculty appointment in Princeton EE. From 2005- 2007, she served as Associate Dean for Academic Affairs for the Princeton University School of Engineering and Applied Science. In 2011, she served as Acting Director of Princeton’s Center for Information Technology Policy (CITP).

Martonosi’s research interests are in computer architecture and mobile computing, with particular focus on power-efficient systems. Her work has included the development of the Wattch power modeling tool and the Princeton ZebraNet mobile sensor network project for the design and real-world deployment of zebra tracking collars in Kenya. Her current research focuses on hardware-software interface approaches to manage heterogeneous parallelism and power-performance tradeoffs in systems ranging from smartphones to chip multiprocessors to large-scale data centers.

Martonosi is a Fellow of both IEEE and ACM. In 2010, she received Princeton University’s Graduate Mentoring Award. In 2013, she received NCWIT’s Undergraduate Research Mentoring Award, and the ABI Technical Leadership Award. In addition to many archival publications, Martonosi is an inventor on six granted US patents, and has co-authored a technical reference book on power-aware computer architecture. She serves on the Board of Directors of the Computing Research Association (CRA). Martonosi completed her Ph.D. at Stanford University, and also holds a Master’s degree from Stanford and a bachelor’s degree from Cornell University, all in Electrical Engineering.

In their free time, Martonosi and her husband Kevin Burkman enjoy hiking, bicycling, and travel. In addition, Margaret enjoys swimming and running. As a member of US Masters Swimming, she tries to do a few races a year, either in swim meets or “open water” (lakes/ocean).

Q: What was your path that brought you to this point in your career? What made you choose computer science in the first place? What made you choose an academic career when you started out?

When I started college, I was thinking I’d probably be a psychology major. But I took a CS class during my first semester, just because it seemed like a good thing for everyone to know. As things turned out, I really liked the CS class and planned to major in it instead of Psychology. But after a few more weeks of freshman year, I decided that I most wanted to understand the underlying hardware of how computers were actually built, and in my mind, that meant Electrical Engineering, so that is what I chose as my major. And that’s the major that stuck. Since then, I have always done work right at the hardware-software boundary, in the topic areas that naturally overlap between EE and CS.

In terms of choosing an academic career… Well honestly, it wasn’t my lifelong dream to be a professor. My Dad was a professor, so I had grown up seeing the academic way of life. But while it looked fine for him, I remember I initially chose engineering in part because I saw that you could get really interesting jobs without going to grad school. I did decide to try grad school (at first thinking I might just stay for a master’s degree) and then similarly at the end of my PhD, I figured I might just try academia for a little while to see how it went. And now nearly 20 years later, I am so happy being a professor. I feel very fortunate that all my not-very-decisive choices years ago happened to lead me down a path that has turned out to be a good one for me. I really like the variety of academia, and I really like getting to work with graduate students over several years and see them mature as researchers.

Q: Explain a bit about your technical interests and how your research evolved over time.

I’ve always enjoyed research that sits at the hardware-software boundary – topics like designing hardware modules that are well-tailored to a particular software characteristics, or thinking about how software characteristics might be changed or optimized in the face of some hardware reality. My early work looked a lot at tuning memory performance for cache hierarchies, particularly on some of the early shared-memory multiprocessors. (Back when they were cabinet-sized, rather than single-chip!)

As I looked more and more at adaptive hardware and software, my focus on power efficiency emerged pretty naturally. Namely, as I thought more and more about tailoring hardware to particular software characteristics, I saw that some of the benefits of such tailoring would be in power savings, not just performance improvement. That observation was huge, because it got me to see the “power problem” quite early, and begin working on modeling and optimization techniques in response.

Another recurring theme is that I like doing experimental research with things to be implemented for real where possible, and data to be measured. That has motivated a lot of my work on performance and power tools over the years – I like being able to measure things and visualize the results, so it was natural to build a power simulator like Wattch, or my more recent work on design space exploration for GPUs.

The other evolution worth noting is that I was a “normal” architect for over a decade before I got into any mobile computing. The transition into mobile computing came through my work on issues of CPU power efficiency. While power matters for everything from data centers to mobile devices, in early days the power crunch was felt first on mobile devices so it was a natural place to focus. When I began working on ZebraNet – a GPS-based mobile sensor network to track Zebras in Kenya – the power-efficiency concerns were considerable, but the more surprising topical transition was in my choice to broaden my research portfolio from simply characterizing the power dissipation of a mobile device (e.g., a phone or handheld computer) towards the wider mobile systems topics of designing mobile sensor networks and new collaborative protocols.

The conventional wisdom is that it isn’t particularly strategic to work on two distinct topic areas (architectures & mobile computing) spanning two distinct research communities. The concern is that neither community will ever see the full scope of your work. But the ZebraNet opportunity was just too interesting to pass up – how often would I get to work with a biologist on an engineering challenge (tracking wildlife across large distances in sparsely-populated regions) that was of central scientific importance to them? So I said “yes” to ZebraNet and it brought me into mobile systems research, rather than simply studying the power dissipation of particular devices.

In more recent years, I’ve shifted to studying human mobility, for use either by cellular network designers or by urban planners, public health scientists and others. So my research portfolio is quite broad, but I find that variety really fun and interesting it is a big part of what I enjoy about being a professor.

Q: What challenges have you had to overcome as a woman leader in the field?  What is the most difficult aspect of your career right now?

One challenge we are always facing is that of finding ways to keep things “fresh” in a career. Since I haven’t found myself particularly interested in administrative roles (e.g. department chair) I need to find “second acts”, “third acts” etc to keep professorial life interesting. I don’t want to feel like I am just turning the publication crank over and over again. So far, interesting new research topics and collaborations have come along at key moments to keep things fun and different.

Each new graduate student and each new research topic always teaches me something new, so that adds variety and interest to the job. I also find it a continual challenge to connect with students in the classroom, and to be as effective a teacher as possible. While I continue to improve in teaching year by year, it is an activity in constant need of practice and refinement. Furthermore, as the age difference between me and my students increases, I need to keep adjusting my teaching style and strategies to align with the current generation. (The undergraduates I teach now were just being born around the time I arrived at Princeton!)

Q: Looking back over your entire career, what accomplishments are you most proud of? Looking forward, what might be your next big thing?

My two best known research projects are the Wattch architecture-level power simulator and the ZebraNet mobile sensor network for wildlife tracking. I am proud of both of them. They have had considerable impact on research and products in our field, and as an academic, it is tremendously rewarding when you get to see that. In addition, ZebraNet will always hold a special place for me personally, because it was an unusual and risky project for a computer architect to embark on, and I usually see myself a hopelessly risk-averse person.

It took some guts for me to say “yes” to starting that collaboration and taking the project through to real deployments, and I am proud that I did it. It was unique and rewarding and we learned a lot personally and technically. What else could one want from a project?

Q: How have you been involved in CRA-W? What has this involvement meant to you? 

CRA-W has been so important for me over the years! My first involvement with CRA-W was attending their first Career Mentoring Workshop in 1993. I was a grad student at the end of my PhD, and the workshop gave me so much concrete advice about how to get started as a post-PhD researcher. In fact, CS has a whole cohort of women researchers roughly my age/seniority who attended that workshop 20 years ago and now remember it as being so vital to their success.

Over the years, I have advised a number of students through CRA-W’s Distributed REU (formerly DMP) program. This has always been a great way to get to work with talented women students, and I have tried to keep in touch with many of them afterwards. Some continue to be my friends and collaborators even many years later.

I was on the CRA-W board for roughly 8 years starting in 2005. I recently stepped off of it for a little break, but hope to return to it in a few years. Being on the CRA-W board let me see (and take part in!) the large amount of work, planning, and fundraising that goes on behind the scenes to keep CRA-W’s programs so effective and vibrant. I have always really enjoyed working with the great people on the CRA-W board; it’s been a great chance to meet people across a wide range of computing disciplines, and to learn from women at all different career and life stages.

Q: How do you balance work and family? Do you have any advice about the challenges for 2-career couples?

In many ways, I think I have experienced almost the best-case scenario when it comes to dual-career and work/life issues, and yes, I do know how lucky that makes me! I met my husband Kevin after I arrived at Princeton, and we never had to do one of those 2-person large-scale job searches that people find so stressful. Rather, we have had different points along the way where we have adjusted to changes, such as Kevin going back to school, or finding a new job. A lot of the good outcomes have been greatly eased by my husband’s flexibility and supportiveness; I am really lucky and grateful. The other thing is that we both really like where we live, so it is not so hard to agree to stay put for now. Perhaps we will pick up our stakes and move at some point, but so far we have found enough opportunities nearby to keep both of our careers interesting.

Kevin and I don’t have children, so the work/family balancing issues are different for us than if we had them. But I think it is important to note that, while being childless makes some aspects different, the fundamental balancing act is still the same: figure out what your priorities in life are, and then allocate your time and energy accordingly. For example, spending time with family has been important to both of us, and we have always tried to make sure we make time for visits with my parents and siblings even when they aren’t nearby. Likewise, Kevin has had periods where he spent a lot of time taking care of his elderly Mom. You don’t know what the future will bring, but maintaining good connections with family and friends is a priority that is worth a lot of time investment.

I feel like in some ways, the work/life balance has gotten a lot easier over the years, now that internet connectivity is so good. I can remember some crazy paper deadlines from 15-20 years ago where I’d go back to campus at 2 am because new data had just been collected and I wanted to meet with grad students there to discuss it in person. Now we can easily do video conferences or email or chats from all over the world to plan our work, and jointly view data or edit paper files. Sometimes, nothing is as good as a face-to-face meeting, but today’s connectivity means I can be home more of the time and still check in on work a bit where needed. Also, it turns out that in the crazy stress leading up to a paper deadline, I am a much happier and nicer adviser if you let me work from home sitting on the sofa in my pajamas!

Q: What activities do you pursue outside of work?

I’ve always enjoyed swimming, and so about 8 years ago with some encouragement from my sister, I joined a local masters swim team. Masters swimming is basically “swim team for grownups” meaning we do coached workouts together, with a bit more intensity and structure than just swimming laps. Thanks to restarting swimming, I lost about 25 pounds, and got motivated to train for (and complete!) events including a bunch of different swim meets and races, a half-ironman triathlon and a couple marathons. More importantly, swimming and my other non-work activities have given me a very broad circle of friends who have a lot of different perspectives, backgrounds and viewpoints – it’s great to be pulled outside the academic/techy world and be able to have conversations with a much more diverse set of people. At the end of a long workday, it is great to head to the pool for a real change of gears.

Q: Do you have any advice for women at any stage of their careers?

Find the joy each day, in where you are now. It is easy to be in a state of constant deferral – to keep saying that you will be happier at some point in the future: after the next deadline, or after you complete your degree, or after some promotion or tenure event, etc. But to me it feels much better to reflect on whatever was the best thing that happened today instead of always deferring joy until some future milestone. Not that you can’t seek to improve your situation, but that even while you are improving things for yourself, it is always possible to reflect on a particular day and find some good in it. Somehow, contentment gets a bad rap these days; some people are so worried about lowering their standards or “settling” that they never allow themselves to feel contentment. That ability to find contentment now while still pushing forwards towards some future goals-that is the ability I wish I had developed and appreciated a lot earlier in my life.