Depends. Right now they're in two different spaces. One still hanging on to graphics/gaming, the other coming from dense compute space. We'll see how long it takes them to converge.
And (as one example), CUDA happily runs on any gamer "consumer" GPU.