I am looking for some things to experiment on and research and maybe try to find a solution. If you guys have any idea about common problems or challenges that haven't been solved until today when to comes to graphics programming, shading, rendering please post here, i would appreciate it.

Thanks in advance

I know a lot of R+D focuses on using consumer grade GPGPU computing for existing massively parallel problem spaces, such as weather forecasting, etc, which would traditionally require super computer clusters on the scale of Los Alamos National Labs. Perhaps you could identify some existing problem like that and figure out how to use GPGPU computing to approximate reasonable solutions with less silicon?

5 minutes ago, Steve_Segreto said:

I know a lot of R+D focuses on using consumer grade GPGPU computing for existing massively parallel problem spaces, such as weather forecasting, etc, which would traditionally require super computer clusters on the scale of Los Alamos National Labs. Perhaps you could identify some existing problem like that and figure out how to use GPGPU computing to approximate reasonable solutions with less silicon?

Sorry i didn't mention at first, but i am talking about real-time rendering focused on game development and not on parallel processing using GPUs. For instance ambient occlusion , volumetric lighting , anti-aliasing. Problems focused on graphics and not on processing problems(irrelevant to computer graphics) which can be solved parallel in GPGPUs with CUDA for example. Also i want to work on low level programming and not on hardware optimization 

This is a good place to start  

https://openproblems.realtimerendering.com/

9 hours ago, Hodgman said:

This is a good place to start  

https://openproblems.realtimerendering.com/

 

I have already checked that out. I am thinking of working and researching on the new RTX technology in ray tracing. Do you think its worth it and has potential?