After implementing the fps display, the performance dropped really hard, because we are drawing the UI on CPU/RAM and then we also have to copy the whole picture into the graphic card RAM. It is a really time-consuming process and because of that I want to regulate it.
The first thought was just to do the calculations directly with the GPU and using the graphic card RAM.
The only programmable interface to the GPU are the shaders and, especially the fragment shader, is made to support such calculations.
I don't know if it will work, but I hope so. A good reference to try out with shaders would be "ShaderToy" (just Google it)
After implementing the fps display, the performance dropped really hard, because we are drawing the UI on CPU/RAM and then we also have to copy the whole picture into the graphic card RAM. It is a really time-consuming process and because of that I want to regulate it.
The first thought was just to do the calculations directly with the GPU and using the graphic card RAM.
The only programmable interface to the GPU are the shaders and, especially the fragment shader, is made to support such calculations.
I don't know if it will work, but I hope so. A good reference to try out with shaders would be "ShaderToy" (just Google it)
If you have another ideas, please let me know!