Ask HN: Getting involved in scientific research computing
A few days ago, a coworker of mine was talking about how his daughter is getting involved in making hardware that can emulate the human brain. That sounded like a really fun project, and it got me thinking a bit about my current career path, which I'd say is a standard large corporate developer. I basically build "text boxes over data." The money is good, especially as I am the sole provider for my family, and the environment is tolerable, but it does get boring at time. Lending my software development skills to scientific research sounds both fun, and something thats aligns with my desire to build software for the advancement of humanity (lofty goal).
Does anyone have an experience or advice on get involved in this type of work? I would probably just be on a part time basis for now, but things could change in the future.
I used to sysadmin for a place doing high performance computing. They were working on a different problem domain than neurology, but still required loads of computational power.
The tech they used:
* GPGPU programming (CUDA, etc)
* Fortran (old code base dated back to the 80s)
* C/C++
* Matlab
* MPI & OpenMP
There was a little bit python but mostly from the newer grad students.
Scientific Linux (https://www.scientificlinux.org/) provides a pretty good base OS though not everyone uses it. Some places use RHEL or CentOS instead. For the GPGPU things the developers used Ubuntu because at the time it was the easiest way to get CUDA running. I have no idea if that's changed.
I've gone down this path. If you are a decent engineer, go speak with academics at your local university. I spent a lot of time with the bio/comp-bio folks, and they can always use collaborators.