Fair point, my bad.No where in my post do I use the word disappointing. From a gaming view, 3x the transistors and “2 node shrinks” for a 50% performance uplift just doesn't seem particularly impressive in terms of architecture.
Fair point, my bad.No where in my post do I use the word disappointing. From a gaming view, 3x the transistors and “2 node shrinks” for a 50% performance uplift just doesn't seem particularly impressive in terms of architecture.
No where in my post do I use the word disappointing. From a gaming view, 3x the transistors and “2 node shrinks” for a 50% performance uplift just doesn't seem particularly impressive in terms of architecture.
The first comparison demonstrates a 140% increase from DLSS reconstruction and than a 50% increase from frame generation.Also 3x the flops. The benchmarks shown so far are a little suspect though. Even the DF teaser numbers don’t make much sense on the surface.
Does frame generation increase performance by 50%?
View attachment 6998
Or by 125%
View attachment 6999
If it’s the former then the raw performance increase would be closer to 100% which is great.
Also 3x the flops. The benchmarks shown so far are a little suspect though. Even the DF teaser numbers don’t make much sense on the surface.
Does frame generation increase performance by 50%?
Or by 125%
If it’s the former then the raw performance increase would be closer to 100% which is great.
50% is disappointing now? Not impressive for gaming?
Hmm, I should go to bed. :[
Faster than what?Probably 50~% faster when looking at a large game performance summary.
Than the Ampere model that would be its predecessor. 4090 vs 3090ti for example.Faster than what?
It's at least 50% faster in badly scaling on Ampere games according to Nvidia.Than the Ampere model that would be its predecessor. 4090 vs 3090ti for example.
I don't believe Nvidia purposely chose titles that showed performance scaling below what is typical. When has performance scaling when averaged over many titles by a 3rd party outlet ever not been lower than the highlighted titles shown in an IHV launch event?It's at least 50% faster in badly scaling on Ampere games according to Nvidia.
You expect it to scale worse than that in other titles?
Note the point of reference, 4090 DLSS2 to 4090 DLSS3 or 3090 DLSS2 to 4090 DLSS3.
EDIT: What techuse pointed out too.
Because it makes the price comparison to prior GPUs look more favorable.Dumb question, but why is nVidia calling the 4070 the "12GB 4080"?
No, seriously. Why?Because it makes the price comparison to prior GPUs look more favorable.
No, seriously. Why?
Many consumers won't know these things though.It's just so f-ing blatant when you look at the memory bandwidth and the core counts on them though, very sleazy.
I actually feel it's like this:Dumb question, but why is nVidia calling the 4070 the "12GB 4080"?
It's just so f-ing blatant when you look at the memory bandwidth and the core counts on them though, very sleazy.