Hi guys, so maybe this is not a very popular feature but, for my setup, which all the cuda device have very different amount of VRAM avaliable: two 750TI,2G vram ; one gtx1080Ti,11G vram ; and an cpu with 32GB of ram that gets used as a cuda device.
So I think the most efficient way to utilize all 4 cuda device is rendering one frame per CUDA device in a sequence right?
thanks guys
So I think the most efficient way to utilize all 4 cuda device is rendering one frame per CUDA device in a sequence right?
thanks guys
Comment