Check this out,the guy has an 3080 and is testing it!
2 Likes
Yes, and if the 60-80% improvement in even classic rasterization performance is anything to go by, then this is a beast. Also, you can add another 20% on top of that for the 3090 plus not being memory capped, even in 4K VR and VRAM intense applications.
Availability for the 3090 is crap though and the first prices for the 3rd party cards are insane (1800 Euros for the Asus Strix 3090 for example - meaning a FE card plus waterblock is still cheaper than a 3rd party air-cooled card, insanity).
I still wonder today whether DLSS was activated on the 2080ti or not. 
Do you know what the availability is? How many?
I’ll buy the 3090 on the 24th Sep if i can!
Sure was. If a game or benchmark supports DLSS, it is activated on both, the 30 series is just plain better at it.
But even if you just go by raw theoretical performance from compute units, clockspeeds etc, the 30 series just destroys Turing. I just wish they would give me a 20GB 3080, the 3090 prices are just stupid. The missus tries to convince me to get one, she probably hopes to score my 2080Ti for free then.
I am not so sure about that.
at least in the Nividia comparsion, 30xx / 2080ti / 1080ti DLSS was not activated on the 2080, only on the 30xx cards.
How do you know? 20202020
The Performance Graph
1080ti 100%
2080ti 140%
3080 280%
With DLSS in compatible Games, the 2080ti have up to 80% more power . (100FPS vs. 180FPS)
So it must be
2080ti 250%
I think
That very much depends on resolution, game, DLSS 1.0 or 2.0. I wouldn’t go that far to outright just claim DLSS not being used at all on the 20 series due to this.
It is probably wise to wait for reviews. The keynotes are not exactly the best place to get performance numbers.
1 Like
40% over the 1080ti are the Standard value without DLSS
But hey, will wait for real user Bench’s before I look if I buy
Not “real bench”, but the video further above in this thread should give you a rough idea what to expect. And that is a steep improvement over previous gen. Doom Eternal is almost double performance for example (2080 vs 3080).
1 Like
yes, but I am interest for VR.
The 2080ti has round 10% more FPS then the 1080ti in VR. How much has the 30xx? Again 10% or 20?
For Screen Gaming, I have a Ultrwide 100Hz Monitor, the 2080ti are enough for this.
Again, a generalization. In what game? At which settings? In such a scenario, you are likely bottlenecked anyway elsewhere, because the raw performance of the 2080Ti is quite a bit more than just 10% higher than the 1080Ti.
Just relax and wait for real world data, it won’t be twice as fast, but the generational leap will be significantly higher than between Turing and Pascal.
The difference between 10xx and 20xx in Vr is not a generalization, but largely a known fact.
I had 1080, 1080ti, 2080 and now the ti and also the direct comparison.
but yes we will wait and see.
“in VR” in itself is a generalization.
There are so many factors at play, that just breaking something down to a single percentage number is oversimplification to the max, at which point it is getting dangerously close to silly hearsay like “Ryzen is smoother in games” (spoiler alert: it isn’t). People say a lot of stupid crap on the internet, even if it is “largely known facts”. Simply because most people are doing a really terrible job differentiating between objective facts and their opinions about them.
The only generalized answer that is always right is “it depends”.
Comparing in same games, same situations, same systems
And the difference between the two cards will differ from game to game. And then you take the average of all the differences to come up with the super oversimplified compound score percentage (patent pending)? If so, why the average? Why not the median?
let us discuss in german, in the sc forum
Sure, always happy to.
1 Like