Faces of HPC: Martin Rüfenacht
Martin Rüfenacht is a Ph.D student in the School of Informatics at the University of Edinburgh. Having completed an MPhys in computational physics, but taking on HPC courses during his final year, Martin realised moved into studying how to improve the performance of MPI.
Martin Rüfenacht is a doctoral candidate n the School of Informatics at the University of Edinburgh working on improving the performance of MPI by using RDMA and is passionate about promoting diversity and making HPC useful to those who need it.
Tell us a bit about yourself.
I’m originally from Zurich and studied at the Zurich International School before coming to Edinburgh to study for my undergraduate degree. I started studying Physics, but switched to computational physics after just one semester, once I realised it was a lot more fun. I’m now studying for a Ph.D as part of the School of Informatics and EPCC Pervasive Parallelism Centre for Doctoral Training at the University of Edinburgh.
During my undergraduate degree I was discussing with my Personal Tutor some of the HPC courses available in to me as the School of Physics is associated with EPCC (the Edinburgh supercomputing centre) and decided to take these courses rather than Physics courses. This was the first time I’d heard of HPC. I’ve always been interested in performance programming, and HPC just became the natural path for me to follow: I’m all about learning how to get something to run faster. I love research, and I realised that even in R&D in business a postgraduate degree is necessary, so a Ph.D in HPC seemed the next best step.
My Ph.D is looking at RDMA in MPI. I like getting the best performance out of machines (hence why I’m in HPC). By using RDMA in MPI you have the chance to make gains not otherwise possible with abstractions.
As part of this project we want to celebrate the diversity of HPC, in particular to promote equality across the nine “protected” characteristics of the UK Equality Act, which are replicated in world-wide equality legislation. Do you feel an affiliation with this matter, and if so how has this interacted with or impacted your job in the HPC community?’
I’m lucky enough that I haven’t seen much discrimination, but then I’m not necessarily at the level where I would expect to see it as a ‘white male’. I believe that the promotion of equality is very important and as such I have attended some of the events organised by Women in HPC. I believe that there are two main reasons why we need to improve equality. Firstly it is immoral if people don’t have opportunities just because of their skin colour or gender. Secondly: if we leave lots of people out of something how are we going to get the best people for a job? I think ensuring that everyone has the equal opportunity to succeed irrespective of their culture, background, gender or any other characteristic is essential.
Is there something about you that’s given you a unique or creative approach to what you do?
I’m willing to abandon all current known ways of thinking of things before I get started. I am happy to throw out the MPI specification if it is not good. I like challenging current ideas and coming up with completely new and novel ideas!
Were there any challenges when you first entered the field? How have you overcome these, or do they continue to challenge you?
I have never been a ‘proper computer scientist. I’ve never had official computer science or software education, instead I taught myself coding while at school and the rest I have picked up. I find that the lack of a formal knowledge of computer science means I am sometimes a bit behind and realise that something was already dismissed 50 years ago, but I’m sure I am not alone. It may be true for other too, but certainly not for everyone.
What is the best thing about working in HPC?
I get to play with high-end machines and all levels (from hardware to abstract algorithms), not just a particular part of it. Its just all fun!
If there’s one thing about HPC you could change, what would it be?
I kind of like it all! I like the HPC community more than the computer science/informatics academic environment. I think because people come from so many backgrounds it is less elitist. I think HPC researchers are looking to help other people and fit around their needs rather than telling everyone the best way to do it is this new way I came up with, regardless of how useful/practical it is to the user.
It would be nice for HPC to be recognised that it is an important part of science. In my background in physics I often felt that computational methods were not respected, that we weren’t either theorists or experimentalists but somehow neither as good as either.
What’s next for you in HPC – where does your career lead you?
Assuming I finish my Ph.D (!) I hope to move to a company, possibly academia, but I’m not certain at this time.
Martin was interviewed by Toni Collis in February 2017.