Going on estimated performance figures not really, AMD have a history of managing to get benchmark results that no third party reviewers manage to get and when asked about it the just shrug their shoulders.
I'm always hesitant to call it based on first party data and estimates so I'm reserving judgment but it's not looking good. It's not looking like we'll be getting the previous generations next model up the stack increase in performance for a similar price to the model under it (we won't be getting 6900 XT performance for the price of a 6800 XT)
Only because they didn't bother releasing anything higher up than the $400 5700 XT.
Exactly. And it's not even a price cut IMO as like i said you can pick up a 6800 XT for near enough the same price.
The 7800 XT should, historically speaking, be called a 7700 XT.
I think you are trying to double dip on generational improvement.
Yes, AMD basically shifted the old stuff down a "tier" in price because they had too many and why would anyone buy a $650 6800XT when new stuff that performs just as well is going to sell for $500?
Dropping the 6800XT's price by as much as they did is effectively generational improvement in performance per dollar.
Complaining that the 7800XT is only offering performance that "has been available at that price for a while..." is basically complaining that they didn't double-up on value improvement this gen.
$500 buys more performance now than it did last gen.
Two years from now, I want $500 to be able to buy another 20-30% improvement. I don't care if they sell left over 7900XTX's for $500 or if they offer something with a new number on the box that gives the performance increase. And I certainly won't complain if they do both.
I expect manufacturers to make my money go faster, I don't care about the names they scribble on the box or the status of "next gen" vs "oh that's old stuff".
Does X amount of money buy more performance now than before. If so, how much more?
For me, 30% seems like good progress. 20% is okay-ish.