The crazy thing about all this is even before I did my research, I knew Nvidia was up to something scummy, pretty much like the 3.5GB GTX 970. First off that 4080 12 GB is not even a 4080, it's not like you are getting the same number of Cuda Cores as the 16GB, they pretend it's two sku's of the same card, but memory is not the only differentiator, the 12GB is cut back, that's scummy...
Second, the 2X increase over last gen is in Flight Simulator and that's with DLSS 3.0 in (performance mode) and I hope to Crom it's not versus Quality DLSS from last gen. Whatever increases to their RT cores software or hardware is a given, but when you are talking 70 to 90TF, we need to see pure rasterization at native resolutions, we need to see the bruteforce power at 450 watts min. The media should be calling this out just as they had issue when AMD put their RDNA 2 tests out in select games and even that was a time when there was no need, but they went in anyway.
So what has NV shown, 3x in a portal rehash and 4x in Racer X, a tech demo, which has no validity as a game you play. The media has not questioned a thing. All I heard today was that Nvidia brought it and AMD is in trouble. Some would even suggest that AMD should give 32Gb cards at $500 to compete. All this power consumption at 4nm and partner cards have not even shown their cards yet, which will probably want to push some more power into these cards.
Also, as I said in my first post in this thread. I knew this was all DLSS stats (in performance mode). No one cautions, wait for actual benchmarks. It's Nvidia afterall, they get the benefit of the doubt even if they don't deserve it. I say Listen to the media's tone on the 12th of October and then on November the 3rd.....