I have to agree with F34R. Why would AMD Radeon be thinking big after Navi? Why weren't they thinking big with Fiji and Vega and now Navi? Were/Are these just placeholders for when they do go 'big', as if going big is nothing more than an active decision to knuckle down and get to it? Navi might be a good product if it can deliver GTX 1080 performance at a low price, but it'll be late and still leave the high-end market vacant.
I haven't seen a truly great Radeon products in years. I thought that would change with Vega, but it didn't—it actually worsened. Then I thought maybe Navi with the Infinity Fabric might change things. Now we're told there won't be a high-end Navi component and it may not use the IF technology, leaving Radeon out of the high-end space until 2020 or beyond. With all that market loss comes less income for further R&D, and the situation worsens.
Zen has been a huge success both because it's a great architecture and because Intel have done nothing but sit on their hands. If Intel had moved to six and eight cores (as they've now done in response to Ryzen) with Skylake 2-3 years ago, Ryzen would have entered the market in a highly competitive space and been a bit of a let down. But in the GPU space, Nvidia haven't been sitting on their hands. Kepler was good, Maxwell was better, and Pascal was better again. So with each recession from AMD in the CPU space, Intel did nothing. But with each recession from Radeon in the GPU space, Nvidia did a lot. If Nvidia's next architecture is nothing to write home about, maybe AMD will have a chance to claw back a tiny amount of market share, but I think it's going to be many years before Radeon is a widely respected brand again.