i agree, it does need more VRAM to output faster, but im more excited about upcoming videos that maintain coherency like a proper human made video, then add audio synthesis to it and we all can implement our ideas and create amazing things. even if the render takes time, still amazing improvement to have.