I know it might sound vague but for the sake of being brief, if you work in scientific computing or high performance computing, how did you find your job? Do you work for a lab/government? I am asking because I want to know what the process is for getting these jobs and whether it can be made more efficient.
I'm currently a fully funded graduate student 'interning' for the USDA and working on (mainly) agriculture modeling software. I found this job as an undergrad through an e-mail system set up at my school for potential employers to reach out to the CS undergrads/grad students for temporary or long-term work. This system can get annoying as even after you have a job, you still get spammed with mail from people looking to hire CS students to do anything from working on their projects to teaching them basic computer skills.
Another note is there are (at least where I work), potential for HPC personnel and they aren't even hiring for these spots. For instance, I was recently talking to a programmer who works with population models. He has no parallel programming experience so he writes his models sequentially and runs them for months and months at a time. If they had a parallel specialist of some sort (even on a consulting basis), their research would be greatly sped up.
I'm a post-doc at a university. I asked around at conferences, and emailed people whose work I liked. I'd be surprised if the process is very HPC-specific; you'd be trying to make scientific job-hunting more efficient. But please do that if you can :D
I work in a government lab and I got my job in a highly conventional manner. I was working on a PhD in a related field, saw a job posting that looked interesting a couple months before my defense, and applied. It all went very smoothly.
Another note is there are (at least where I work), potential for HPC personnel and they aren't even hiring for these spots. For instance, I was recently talking to a programmer who works with population models. He has no parallel programming experience so he writes his models sequentially and runs them for months and months at a time. If they had a parallel specialist of some sort (even on a consulting basis), their research would be greatly sped up.