I think you're missing the point. It's not about the VRAM. The 3090 can be at
most 20% faster than a 3080, and will likely be less, since even 20% would require perfect scaling of on-paper specs, which never happens. Let's look at the 2080 vs the 2080 Ti:
48% more CUDA cores
38% more ROPs
30% more SMs
352-bit vs 256-bit memory bus
Actual performance gain: 28% on average (at 4K, much less below)
https://www.techpowerup.com/review/nvidia-geforce-rtx-2080-founders-edition/33.html
The 3090 has over the 3080:
21% more CUDA cores
17% more ROPs
21% more SMs
384-bit vs 320-bit memory bus
Actual performance gain: ???
The 3080 is relatively MUCH closer to the 3090 than the 2080 was to the 2080 Ti, and there was only a 30%-ish performance gap between those two. It's going to be a lot closer between the 3080 and 3090, because GPUs aren't magic and the specs are known. There's almost zero room for a 3080 Ti in the performance stack even as it stands, unless people are prepared to pay 50% more for 10% more performance. Then you add the existence of the 3080 20GB on top of that and you don't even the VRAM factor to differentiate it. That's the relevance VRAM has to the discussion - marketing. You could maybe sell a 10% faster 3080 Ti on the double VRAM. Maybe. But you can't do that if there's a 20GB 3080, which AIBs are saying privately that there is.
