Frank Austin Nothaft

Going Back to Grad School

Today, graduate school begins.

About a year and a month ago, I decided to apply to do a PhD in computer science. At the time, I had been working at Broadcom for a year after completing my undergraduate degree. During my undergraduate career, I was fortunate to do research with several different faculty in several different areas of electrical engineering and computer science. I also interned in industry two summers–one summer each at SAIC and NVIDIA.

Going into the senior year of my undergraduate degree, I knew that I liked research, but I also had greatly enjoyed my time working in industry. My girlfriend at the time (who is now my wife!), was moving down to Southern California to start graduate school at USC. At the time, none of the schools in the Greater LA area were doing research that lined up with my interests. This pushed me into industry, and I found a job working at Broadcom in design automation and verification.

At the time, many of my friends and advisors warned me that going off to industry likely was the death knell for my academic ambitions. Many people pointed out that it is very difficult to go into industry, make a stable salary, make many friends, and to then give that up and go back to school. In some ways, I chose to take this to heart. I told myself that a PhD is a very large investment of time and energy, and I would use my time in industry to discover whether I was really more interested in research, or implementation. I knew at that point that if I decided to go back to school later, I would know exactly what I was getting myself into, and what I was giving up!

Two years later, I know that this was the correct decision:

  • First and foremost, a 24 year old is much more mature than a 22 year old. I’ve seen more technical problems, and I know more about myself, and what I want out of my career.
  • For an engineer, industry is the school of hard knocks: industry taught me to be a good engineer, as opposed to being a good engineering student. Customers demand perfection, and there is are several magnitudes of complexity that separate a good class project and a good product.
  • My work at Broadcom was about 75% ASIC implementation, and 25% CAD research. This mix taught me two critically important things:
  • Semiconductors are a booming field, but it’s not clear how much longer they’ll continue to boom due to the demise of Moore’s Law, and difficulties scaling to smaller technology nodes. On top of that, the big areas of growth (RF, bio, MEMS) are not areas that I am strong in.
  • I like working on products (and tend to get fiercely attached to them!), but my real passion is research. If I had to work from 7 AM to 2 AM just so I could squeeze several hours of research in, that was not an obstacle.
  • One last thought for anyone reading this: industry may ebb and flow, but the opportunity to go back to grad school is always there. However, a caveat exists: this opportunity will diminish if you do not keep in contact with the academics who you have previously worked with.

Long story short, it’s been an interesting ride, but I do not regret any of it. I had the opportunity to work on a fantastic team of individuals, working on cutting edge products. Two years later, I am headed back to grad school. In many ways, I’ve been able to have my pie, and eat it too.