I am working on an installation which has to output a 7680X1080 resolution on 4 tv screens.
The content is some particle and fluid simulation controlled by 2 kinect azure sensors tied in a line across the screens.
The rig that I have for this is an i9 2.8 ghz with 32 GB RAM and a 2080 Super 8GB.
7680X1080 is actually a 4k output but the hardware team connected 4 DP/HDMI cables in the GPU and I am limited to outputting a resolution of 7680x1080 in the final window which I find is really power consuming and pushing the GPU to 98 temp when player in range playing with the fluid simulation.
The fluid simulation is rendering a 7680x1080 content because of the layout of kinect and the body skeleton that I am rendering with an orthographic camera (can I have a different approach on this?).
So my question is if asking for a better distribution of pixels in the final window like a single 4k output would improve a lot my performance and maybe asking for a more powerful GPU like a 3xxx series or a Quadro or Titan? Also do you recommend doubling RAM memory?
Also I need to keep in mind this thread which says that azure sdk have some problems with 3xxx series gpu.
Any help would be really appreciated. Thank you.