i did this at .. i think 32. i'm glad i did, i did't go to undergrad, so I got to learn algorithms and theory and some other things. completely unprepared for

i did this at .. i think 32. i'm glad i did, i did't go to undergrad, so I got to learn algorithms and theory and some other things. completely unprepared for the social reality of being a student at an associate professors age. its also pretty difficult looking at some of the material from the perspective of a commercial developer. in your case it depends alot on whether or not you want to work on the physics side or the code side. if its the former your choice is pretty clear, if its the latter I think you'd be surprised at how starved HPC is for talent of any kind. if you're already in the US lots of labs hire pretty junior people and there is a fair amount of room you can grow just on your own inertia. there are also alot of fairly junior positions in industry in sales engineering and related roles doing benchmarks. if you show an interest there is room there also to grow into at least maintenance development. again, if you are in the US - LBNL/NERSC/UCB is a great place to do work in adaptive grids. see if you can meaningfully get employed in some support role and pivot. they are also really nice people and would love to let you grow into a role. another advantage of working as staff, even junior staff at a university or associated lab is that you get access to courses as a non-matriculated student. that way if you do want to take grad level scientific computing or analysis or discrete time domain pde solutions everyone is perfectly happy for you to go to class free from 2-3 MWF anyways, if you're sure that fluids is your future, maybe just try to cut to the chase a little. hope you are ok with fortran though.