This is the CPU rendered (13/128 samples)
This is the GPU rendered image
I think the first has better shadows but the 2nd is just brighter.
I've been modeling for a while and was just playing around with 2.8.
Also, should it be text, video or both?