Well, to have a direct desktop version of the PS4 Neo chip kinda makes sense, especially with all the ports we get these days. I imagine the reasoning is even less work during porting.

I ain't a fanboy, I don't care what card I buy. I buy whatever is best for me at that time. So, I really hope AMD are just messing around, and drop a bomb on us that involve insane high-end (and above) card performance with even more insane value.

...AMD is spinning their inability to compete at the highest end. The end.

Did not know that the mention of "mainstream" VR is not pure spin? Just being a parrot; needs cracker. The whole point, it's their spin on the cards. Generating hype by saying nothing more than the price for 390 series replacements is less than the 390's are currently. That and the power consumption drop. Every piece of marketing so far has pointed to this thus far; ignoring rumors. In other words, they have said nothing credible about the cards except for relative price. Until AMD makes an official announcement, everything is rumor and should be treated like all marketing; hot air, puff of smoke. Everything else is arguing which grain of sand has more sides.

That's definitely not happening around the time of 1080/1070 release.

It isn't so bad if it's 390x performance (probably will exceed it in some cases) at 380x prices, with a low power envelope they can use those same GPUs for mobile, a market which they had been outclassed in recent years... Anyway, Vega will be the GP104 competitor, and we can safely assume it will have ~4096 SPs like fiji, so Polaris' clocks will give us a rough estimate of what we can expect from vega

If the 480 can come close to the 390X for around the price of a 380 as a earlier article suggested the card would still be a steal for non 4k users. I do think AMD is making a big mistake for coming out with lower/midrange cards while Nvidia is coming out with 1070/1080's.

I think they would do just fine due to the fact you have alot of people who just dont want to drop 500+ on a video card just to find out a new better one is already on its way and my guess for the reason to not releasing there highend gpu's is due to zen coming soon.

I think they would do just fine due to the fact you have alot of people who just dont want to drop 500+ on a video card just to find out a new better one is already on its way and my guess for the reason to not releasing there highend gpu's is due to zen coming soon.

The problem is not the 500+ GPU, the problem for AMD is the $389 1070. The performance/price ratio between the 1070 and Polaris 10 is what will matter in the end. If you give an extra $100 but you get 50% more performance, then AMD has a problem.

Helluva lot of people can't or won't always drop 100 dollars more. There are few places in PC technology where you can't always add a little more to get a better part. That 100 dollars would be already quite a lot more at that price range. It's not like between 900 and 1000 dollars. Time was just right for the Maxwell lines of cards, and that's why 970 sold like hotcakes. In fact I was very close to buying one to replace my previous 560Ti. The generations between the 500 and 900 series didn't catch my attention at all. But is the same true for Pascal? Who knows.

Helluva lot of people can't or won't always drop 100 dollars more..

In what world 150w TDP part is a mobile chip? If that is the case then frikkin GTX1080 has a mobile chip inside too with its 175w TDP.

So if i say the next "smaller" gpu is 290 (vs 1070=390$), I would have to do crossfire/sli to get the same gain in fps. 2x290 or even 250$, means im paying 150-200 more than compared to running the next faster single gpu. Not sure where going with the more expensive card doesn't make sense. One major reason i always recommended to spend a bit more to upgrade, rather than buying another (of the smaller card) to do crossfire/sli later on. Unless we're talking about a top card like 980ti or tiran, where i can only gain fps by adding another card..