....NVLink combines the memory of each card to create a single, larger memory pool.
Petersen explained that this would not be the case for GeForce RTX cards. The NVLink interface would allow such a use case, but developers would need to build their software around that function....
So it would not be the case, yet it would allow such use case... such which one is it? Kinda contradicts itself.
And when Petersen says, it does not "automagically double frame buffer, it is more nuanced than that" - i guess on Quadros and Teslas it is magic, then? And why, as a consumer, i should care how its achieved in the first place? All that matters to me is, that is works.