Glsl

Troubleshooting General Help
1

Hi,

at my knowledge, the older GPUs use 24 or 32 bit floats and the newer chips support 128 bit floats. With glsl, do I use a standard float with a specified size or the float type of my graphic adapter?



Best,

Andreas

2

There's no way to change the precision of floating point types in GLSL, so how exactly can you choose which float type to use?

3

Does it mean, if my graphic adapter supports 32bit floats, glsl float's precision has also 32 bits and if my hardware supports 128bits, glsl float's precision has also 128bits?

4

Hi,

I just read here: http://www.maschbau.fh-muenster.de/~boesche/www/03/opengl/opengl.pdf, 128 bits mean a vector consisting of 4 32bit floats. If the vector or the register has 96bits, then you have 4 24bit floats.



Best,

Andreas