The PhD End-Game (Thesis/Viva) Experience

So you start your PhD in HEP in Imperial by doing courses to get a grounding in the subject, work out (hopefully) what you actually want to do as a PhD during the rest of the year, spend the better part of the rest of the three years (or four, as it is now) desperately trying to create a cohesive piece of research, then bind the whole lot into a thesis. Easy.

Perhaps the strangest thing for me about writing my thesis was really trying to decide what point I was getting across. During a PhD you always end up working on a lot of different things, some of which you include and some of which have no direct relevance. Then you have a restriction on how long it should be. Mine was 170 pages, a little on the long side, but I actually left out a huge chunk of work on a medical imaging project called I-ImaS we worked on in the Silicon lab. It would have muddled the topic of the thesis, not to mention doubling its size.

In the end I decided that CMS was my primary focus, and the important work was on the off-detector trigger and readout electronics. As a CASE student I was in a slightly unusual position as my PhD was mostly about the hardware design of CMS, including the complexities of design, implementation details and possibilities for future upgrades. I also became heavily involved in designing and commissioning the Global Calorimeter Trigger (GCT), very late in my PhD. While it delayed my submission it also shifted the topics I chose to present.

Approaching things from a hardware perspective left me with an interesting dilemma when it came to the viva. What to revise? Hardware, analysis methods, trigger algorithms or more of the physics? As CMS isn't a running experiment this is more of an open-ended question than usual, and you can't know everything (there are, after all a few thousand people involved in the project for a good reason!). In the end I did a bit of everything, but (and perhaps this is due to the hardware nature of the PhD), I think that it's harder for examiners to relate to someone who's specialised heavily in hardware, as it's a less common trait in particle physics. Working on FPGAs and ASICs in Silicon requires a lot of specialised knowledge and isn't an area that's accessible to most physicists. (Incidentally, I think people often forget how lucky we are at IC to have a Silicon lab and an electronics workshop, not to mention some very dedicated and capable people to work with in both of them).

My external examiner was, (unfortunately for me!) very knowledgeable in FPGAs and the modern technologies used in CMS. I had a couple of very interesting discussions about the intricacies of 8b10b encoding in serial links... but one thing became clear during my viva. Virtually all the discussions were on topics I hadn't expected. This isn't to say that reading around the subject before the exam isn't useful, but it's worth remembering that you're partly there to be examined one what you don't know rather than what you do.

Which brings me to my conclusion on the experience. I think if you work hard, write a good thesis and contribute well to whichever collaboration you work on you'd have to work hard to fail in a viva. Your fate is most likely sealed before you even enter the room. For me I think the most useful thing about the viva was the clarification of what I didn't know. As I'm staying in research this is something I can work on!