Ask HN: PhD in Compiler/Programming Language Theory

3 points by jetti ↗ HN
I've been thinking of getting a PhD focusing on Compilers/Programming Language theory but am not sure if I should do it. I have a MS in Computer Science (but a bachelor's in Economics) but I didn't get a chance to take the Compiler (Advanced) classes but did get to do PL theory. If I were to get the PhD is there a need for people in that field (not in academia)? I know there are a few people that are in Microsoft that have their PhD and are working on the languages such as Don Syme and Mads Torgersen. I just don't know if there are many jobs that would require that kind of expertise.

4 comments

[ 4.4 ms ] story [ 22.0 ms ] thread
I'd suspect that a Phd in computer science would probably have a skill set that is attractive to a fair number of employers. I suspect that the knowledge, skills, and abilities that would be necessary for obtaining a Phd in the area you mention would also be attractive.

I suspect that supply outstrips demand in academia as it does in pretty much every doctoral field. In industry there are people who work on compilers at many companies. In terms of new programming languages, there is and there isn't. There is lots of normal line of business work that benefits or would benefit from domain specific languages. On the other hand, Google's got Brian Kernighan if they need a new major language.

The thing that worries me is that there are somewhat limited compiler jobs and so if I have to wait for one the PhD would hurt me as my specialty would not be something that is every day stuff.
My childhood friend Mike is a Phd organic chemist. His dissertation was on detecting chemical weapons. He went to work for Coke a Cola. I don't think it was due to a strong opinion about high fructose corn syrup. I think it was because his training made him able to learn the aspects of organic chemistry that were relevant to the business of selling sugar water and because he knew his way around a chemical lab and equipment. His Phd also got him an interview and qualified him for the position. That seems pretty typical regarding the vocational aspect of getting a Phd.

An interest in compilers would be valuable for sustaining a person over the span of years it takes to earn a Phd. On the other hand, I don't know would and wouldn't be considered 'a compiler job' by you and who knows what the 'compiler job' market will be in a few years and how it will relate to what tiny piece of compilers you might work on while pursuing the degree.

Anyway, where you go to school and who your advisor is will also affect how relevant your study is to company's that hire people for 'compiler jobs.'

Good luck.

Thanks for your feedback. I never really thought about the skills transferring but looking at it the low level optimization that would occur when researching compilers would give me a leg up on positions that would require that kind of work.