Andrew King, Director of Performance Research

Andrew King completed his PhD in Computer Science in 2009 with a theoretical thesis focused on graph colouring, which describes how networks of objects can be organized. Today, he does experimental research on quantum computers as Director of Performance Research at D-Wave Systems.

Q: To get things kicked off, when and why did you decide to do a PhD? 

I did my masters at the University of Toronto and I got one or two recruiting emails from big companies looking for applicants, but I decided I was not interested in getting a job. I was more interested in continuing to study things that I was interested in. At the time, my brother had recently done an NSERC USRA at McGill over the summer. He told me there was a prof working on graph colouring, which is what I was interested in, and that he’s probably taking students. So, I emailed him, and I joined him at McGill in 2004. He ended up being an excellent supervisor and a great fit for me, so it worked really well. 

Q: What was your experience like establishing your relationship with your supervisor and figuring out how much feedback you needed?

That was a really difficult thing. I’m sure it’s a really difficult thing for a lot of people. My supervisor is definitely more on the informal side, a little bit maybe towards the disorganized, mad scientist, personality type. He’d come up with an idea for something for me to work on and he’d explain it to me…and I wouldn’t know what he was talking about. He’s kind of an intimidating character, so it was a real struggle for me to frequently go back and say, “I don’t know what you’re talking about, I don’t understand any of this”. You kind of feel like you snuck into this position of being his grad student and you don’t want to admit that you don’t know the research area. Admitting your ignorance is something that is an extremely valuable skill in my professional life, and I like to think it’s something that I developed as a grad student. 

Andrew King reflects on the challenge of admitting ignorance when presented with new ideas by his supervisor.

Q: Going forward in the timeline, what was your career outlook or career plan? Were you planning on becoming a professor? 

I had hopes of landing a faculty position, which didn’t happen. I did what was supposed to be a postdoc in the Czech Republic, but I never managed to finish my dissertation before I left, so it ended up kind of being a victory lap on my PhD and I wrote up while I was there and defended when I got back. Then, I did a postdoc at Columbia University, and I did a postdoc at Simon Fraser University, then I landed in industry. 

All the while, I was looking for a faculty position. Towards the later part of that process, I was getting some promising interviews and I actually had a pending decision when I took the job at D-Wave, where I am now. I decided that if I was going to do something other than being at university for the rest of my life, then it was a good time to do it. At the level that I was at, I could have found a faculty position, but I would have had to be a bit less picky about where I was going to live and what kind of school it was. 

Q: Can you tell me a bit about what your job is now? How does it compare to the type of work you were doing as a graduate student?

I work for a company called D-Wave Systems. It’s probably 25% PhDs. Most people that I work with have a PhD.

D-Wave builds a certain type of quantum computer called a quantum annealer. It’s bleeding edge technology, so what I work on is trying to figure out what it does well, why it does it well, why it doesn’t do other things well, and trying to produce research results that showcase its strengths.

The work is much more experimental than what I did as a grad student. Personally, I think that it’s a much better fit for me. I was a bit ill-suited to theoretical work, actually. I was OK at it, but I was never amazing. I think I’m very strong in experimental computer science and, because I have shown an ability to choose projects that are likely to succeed and follow them through, I’m given a lot of freedom at work. 

It’s a very academic environment, but, of course, we don’t have teaching responsibilities, which is amazing. I get a lot of time to focus on the work that I want to do. 

Q: Are there any specific lessons or experiences from your PhD that serve you in your current role?

Actually, one thing that sticks out is that my adviser was a real stickler about writing and communicating technical ideas very clearly. With any draft it was kind of that cliché situation where he sends it back, and sends it back, and sends it back. It’s one of the reasons it took me so long to finish my PhD, but in hindsight, it was very valuable. 

For him, you had to be communicating extremely clearly all the time and choose the structure of every sentence very carefully. Being in an interdisciplinary environment now makes that much more difficult because you need to choose your nomenclature such that you broaden the scope of people who will understand what you’re talking about. 

Q: Can you comment on the differences between the challenges that you had to overcome during your PhD and the challenges you overcome now in your work? 

There’s a really big difference between being stuck in an experimental research project versus a theoretical one. A theoretical one is way more frustrating because you hit a dead end and often you just have to give up and abandon it because you can’t find a way through or because you realize that there is no way through. But with experimental work, in the worst case, you can always blame the person who made the apparatus [laughing]. 

I find that it’s much less frustrating and is even interesting to find incongruities in my research now and ask “What is this problem? Where is it coming from?” It often leads to identification of problems with the quantum hardware that we hadn’t known about before. So that’s rewarding. It’s more rewarding to fail in an experiment than it is in theory. 

Q: Is there something that you know now that you wish you knew when you started your PhD? 

I would consider my story a huge success story for a PhD going into industry because I have a very high level of compatibility between my work and my area of research, and I’m very happy with it. But it took me a long time–I didn’t get the job until I was 33. 

As an academic, you uproot yourself and then hopefully land somewhere. Thirty or forty years ago, you land a faculty position before you’re 30, and then you can have your life. But when you’re drifting around from postdoc to postdoc, moving from city to city every two years, it makes it really difficult to have a life. As you get into your 30s, that becomes more and more problematic. I don’t regret doing a PhD, and I’m really happy where I landed by all the coincidences of life, but I do wish that I had started looking for industry positions a little bit earlier. 

Many thanks to Andrew for sharing his PhD narrative! You can find out more about him on LinkedIn and read some of his recent work published in Nature and on arXiv.



This interview took place in June 2020. Interviews are edited by the TRaCE McGill Editorial team for length and clarity before publication.

Discussion