It's not as if Nvidia is always a better choice. Things turned around a bit when new crop of games (The Last of Us, Hogwarts Legacy, etc) came out, requiring more than 8GB of VRAM. They do work with 8GB, but with lot more stutterging as textures have to be constantly transfered from main RAM. For these games amount of VRAM made a huge difference. At that point in time, AMD had more mid level cards with 12GB (6700) and 16GB (6800) and for considerably less money than equivalent Nvidia, at least in Europe. Nvidia cheapened and stuck to 8GB even for 3070 which was more expensive than 16GB 6080 around here.. This became a big issue and pushed Nvidia to put more VRAM in next generation.

If you actually try and find proof of the 8GB fiasco, there's mostly one source that put together an artificial scenario where they pushed a bottom end card to 4k on newly launched console ports which then got optimized.

NVIDIA said as much, released a 16GB version, and lo-and-behold it's absolutely pointless, even performing worse than the 8GB at times: https://www.tomshardware.com/reviews/nvidia-geforce-rtx-4060...

Most graphics cards are still under 8GB of VRAM, and raw numbers of memory storage mean nothing for performance in isolation.

I believe it's the 128-bit interface hindering performance more than the amount of VRAM.

The framing here is a little unfair, its DLSS that does not support AMD hardware, not the other way around. This is clearly a deliberate decision by Nvidia. Meanwhile AMD/FSR supports both AMD and Nvidia hardware.

The framing is fair unless you expect NVIDIA to ditch their ML based solution for a class of inferior implementations which in turn can run just about anywhere.

DLSS isn't "NVIDIA flavored FSR", it's a more advanced pipeline that requires significantly more architectural alignment. FSR 3 was going to be the closest thing to an equivalent and won't run everywhere FSR 2 does.

DLSS directly leverages the tensor cores of particular NVidia cards.

Whether or not this is just a marketing gimmick is up to you and your own conclusions on the matter, but AMD doesn't seem to have an equivalent hardware tech for this yet. It seems to be taking steps in the right direction with the WMMA instruction for accelerating matrix multiplication in a similar fashion, but performance so far head-to-head seems lacking, and I haven't seen anything definitive out of AMD suggesting this for gaming purposes.