i have implemented application, generating 2 geometries on 2 different gpus using opencl kernels. want render these geometries on 2 different gpus simultaneously , combine final images. how can determine want render first geometry on first gpu , second geometry on second gpu, using opengl api?
well, since system level apis creating opengl contexts lack device enumeration , selection, more or less reliable way create window within boundaries of screen connected desired gpu , create opengl context using hdc of window (while within boundaries of screen and visible force driver); 1×1 borderless window suffices.
if have monitor(s) connected 1 gpu, well, out of luck. sorry.
Comments
Post a Comment