Either way you end copying to the graphics card... Does this mean that a dedicated framebuffer is not nessecarily faster than a CPU, but is only good for offloading calculation power? And the CPU is only faster if you do nothing else but blit a bunch of pixmaps? This in the sense that if you do more than that it would be faster to let the graphocs card handle the blitting because then you don't have to move stuff back and forth between the CPU and the GPU to avoid massive latency?