As I said, AMD and their fake misleading benches.
We know now why the 5xxx series seems so fast in gaming. They used the smart access memory to inflate the results.
Here it is the same, also using heavy overclocking to even match the competition. Imagine Intel or Nvidia doing this, presenting OC cpus or gpus and comparing them with stock versions of the competitors, just lol.
Smart memory is a feature that is available with the right mainboard+CPU combination, so yes, it does not apply to everyone. AMD is using this to entice customers to go all AMD.
As for „heavy overclocking“ - if Rage mode only improves performance by 1-2% I would not call that heavy or even overclocking. It would be interesting to see if and how that affects power consumption, most importantly do cards stay in the TBP budget given by AMD.
As for overclocking, one could argue that Ampere, or at least the 3080 and 3090, is / are already running very close to their limit. If Navi 21 has headroom, that‘s a plus. Note that AIB cards are rumored to offer noticeably higher clock speeds.
One more thing: The 6800XT was also shown vs. the 3080 without Rage mode and smart access memory and numbers looked pretty good. Testing for AMD and nVidia GPU was done on the same platform.
Still, there are many open questions wrt RT performance, their DLSS alternative, power consumption details (what does the number include— everything on the board including USB-C, or just parts, is it a typical or max value..?), have there been updates to Video Core Next....
What we need is an architecture deep dive and of course third party reviews.
But as always, get what you feel is best for you.
I for one feel that if AMD‘s numbers are accurate it means there will finally be a choice from top to bottom in the GPU sector, something we haven‘t had for a long time. And this is good.