DF 5070ti vid is up.
Turns out it really only is 25-30% faster than my 3080 in real use scenarios. Lol where is double performance as seen in some benchmarks yesterday?!
And how much is 5080 faster than 5080ti?
Man what a complete nonsense of a lineup. 750 (1100$ here) card being, 30% faster than 2020 700$ ram. Granted, better DLSS features and 16gb of vram... Again, the price is an issue here.
I would saym, any 4080 (super or not) is a better buy than both of these cards. I don't think that's too far fetched is it? Especially with physx removed too
cmon at least msrp
Real use scenarios? Techpowerup benchmarked 25 games, and at 4K the 5070ti was 51% faster (relative difference) on average. This means that in certain games the difference can be lower or higher than 51%.
In Black Myth Wukong (lumen) the 5070ti is 57% faster than the RTX3080
200% relative difference with path tracing
PT in black myth wukong at 4K native destroys every single GPU, so it's necessary to use both DLSS SR and FG in order to get playable results. I dont know what settings you are using on the RTX3080 in this game, but here's my own results. The RTX5070ti should get similar results, especially when OC'ed.
4K DLSS Performance + FGx2, very high settings, full RT (PT). The game is perfectly responsive at these settings and I think the image quality still looks great. I disabled excessive in-game sharpening and used my own reshade settings. I also forced the Nvidia APP to use the latest DLSS (transformer) settings (I'm loosing 2-6fps, but image is sharper)
Screenshots taken with these settings.
4K DLSS Performance FGx2, very high settings, medium RT
4K DLSS Balance + FGx2, very high settings, medium RT
4K DLSS Performance + FGx2, very high settings, lumen
4K DLSS Quality + FGx2, very high settings, lumen
1440p DLSS Quality + FGx2, very high settings, full RT
Screenshots taken with these settings.
1440p DLSS Quality + FGx2, very high settings, medium RT
1440p DLSS Quality + FGx2, very high, lumen