I'm a regular (non games, I work on enterprise operating systems and databases) developer by day, looking into starting a project where I can apply geometry and fast math related algorithms which leads me straight to graphics programming and specifically programming for games (real time, performance critical, very hardware driven).
I honestly am not very concerned about making an actual game quickly or getting an app out there very soon but I do want to implement code that run super fast, renders well and simulates realistically. I have reasonable skills with C/C++, x86 assembly and python plus a fair history of math.
Here's what I have so far:
1. I've done a basic graphics course at uni (long ago) and implemented some of the simpler algorithms while there.
2. I've set up a nice environment where I can draw stuff at a near pixel level using X windows (I know OpenGL does things better but while I'm learning it feels like too much of an abstraction)
3. I've implemented a couple of algorithms to do rendering, basically some simple shading and drawing algorithms. I've been following the math and for game developers videos here although I've implemented far less of those than I'd like to admit.
Here's where I want to go:
1. Proper hardware acceleration for the algorithms that I've written, ideally using a dedicated graphics processor. How do I access these? I'm reading through Michael Abrash's guide but it seems very dated. Is there something similar out there for more modern processors with nicer hardware features?
2. Is there something out there that covers the latest topics in rendering (like an academic type text) and the research that's going on there?
Ideally good algorithms to use with modern systems.
3. I want to move onto physics too, I was hoping someone could point out some nice references. I've got a few books already and I'm going to start implementing the algorithms but I would like to finish the rendering aspects first.
Finally, if someone here who works within the industry could point out a good area to focus on, that would be super cool.