Hey Rus, isn't Vega 10 the big chip (meant for Fury) and Vega 11 the smaller one (meant for RX 490) ?
Kaduri noted that the naming convention is related to the timeline, not size of the chip. If Vega 10 and 11 come true, the 10 version launches first. It should be a smaller and slower chip than the 11 version. The concerning part is if Vega 10 is meant to be a 1070/1080 competitor, then when is Vega 11 launching? If Vega 11 is also 6-9 months behind GP102, then that would be bad news. One can hope that if AMD is 6-9 months behind with high-end cards that they will either try to beat NV outright or at least shake up the price/performance in a way R9 290 did.
---
32 ROPs and lack of GDDD5X only reinforce the mainstream/performance targeting aspects of RX480. Since this is an R9 380/X replacement, I wouldn't start freaking out about a failed "Fiji" style chip. HardOCP has no clue what they are talking about. If AMD was designing an R9 390/390X replacement, it wouldn't be 232mm2 and have only 256-bit bus and 32 ROPs. It can't be more obvious that HardOCP is just wrong, and yet we see people repeating over and over how RX 480 is a failed 390/390X "successor".
Also, I remember so many people wishing for Polaris 10 to be as fast as a Fury/X/980Ti and I warned against setting unrealistic expectations. Low end/mainstream chip = low end/mainstream price. The hype got out of control.
The concern here is if NV matches or beats 480, AMD will most likely have a gap in the entire $300-700 range until Q1 2017. I expect to read A LOT of biased/NV PR posts about how Polaris for GPUs is AMD's Bulldozer for CPUs (i.e., AMD is officially out of high-end/performance space). AMD's desire to launch Bottom-Up this time seems to have confused most of the Internet and even some review sites...