-
Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.
-
We have made minor adjustments to how the search bar works on ResetEra. You can read about the changes here.
[DigitalFoundry] Nvidia GeForce RTX 3080 Early Look: Ampere Architecture Performance - Hands-On!
- Thread starter dex3108
- Start date
With DLSS 2.0 you won't really have to make the tradeoff, either, if you do end up getting a 4K set. The 2060 super already has pretty good frames at 4K with DLSS on in Control at Ultra settings for example. So these cards will be at LEAST 50% better than that.
Well, I wasn't expecting that much of a leap, and relieved to see that pricing hasn't changed significantly.
This has greatly increased my interest in a desktop build.
This has greatly increased my interest in a desktop build.
Evidence of DF not being objective?
Plenty of outlets get early previews, review codes and are objective.
These numbers are so much higher than I was expecting based on leaks. Crazy. I was planning on air cooling but these numbers make me want to go water again just to squeeze everything I can out of it.
I told you, RT scales with shading performance. These 4xRT rumors from MLID were bullshit and that video demonstrates it perfectly.
Control without DXR has the same performance uplift from the 2080 as Control with DXR.
Control without DXR has the same performance uplift from the 2080 as Control with DXR.
Taking into account that 2080TI is 250 watts and 3080 320 watts, a 28% more, the efficiency gain is negative taking into account the node change.
Yeah that performance is off the scale. I was thinking how close next gen consoles were to high end PC's coming into this gen, and theya re still a great value in terms of performance per dollar, but in terms of just overall performance... nah, not even close to high end. A 3060 is likely going to be much better than a next gen console, nevermind these beasts.
this seriously makes me consider just building a pc. money isnt really an issue but damn. i dont want to be stuck with 30fps for another gen.
I love that they used a Nintendo HDMI cable to test it. XD
Apparently that cable is better than I expected.
Apparently that cable is better than I expected.
Yeah I mean I expected them to be much better spec wise without question than either console but as always we have to let the games show us the difference visually as the generation goes on.
They did a bunch of Nvidia sponsored videos lately, i also think they should be careful with that.
With DLSS I agree, without it there will be a need for a mid-gen upgrade.
Just for kicks (waste of time really) and based on framerates of previous DF SotT benchmarks with a 2080, the third scene was getting in the 42-45 fps on a 2080 at 4k, or ~70 on a 3080.
Showing up for warmups
I'm not upset about it. I'm not judging RDNA2 or AMPERE on first gen let alone first 6months of drivers.
Shame DF had to do that at the start but anyway, this is really impressive and not a bad price. I had to look twice at the TF someone posted here it's just nuts.
Holy...We just reached toy story 5 graphics!
I wonder how the TDP compares to the 2000 series..
So guessing 3060 is 12 TFLOPs at least? I want a laptop that does ray tracing and is 2060 to 2080 specs but with a lower tdp than the 2060. Hopefully they offer a model for that. Might have to be lower than 3060...
So, using some benchmarks online, I believe this chart should be fairly accurate based on the performance % increases Rich was getting from the 3080 compared to his 2080 in the video.
Borderlands 3 (4k/Bad Ass/DX12/4K/TAA)
RTX 2080 - 34.33fps
RTX 2080 ti - 48FPS
RTX 3080 - 61.79fps (180%)
Shadow of the Tomb Raider (4k/Highest/TAA)
RTX 2080 - 47fps
RTX 2080 ti - 61FPS
RTX 3080 - 79.81.59fps (168%)
Battlefield V (4k/Ultra/TAA/DXR OFf)
RTX 2080 - 67fps
RTX 2080 ti - 84FPS
RTX 3080 - 112fps (168%)
Control (4k/High/TAA/DLSS OFf)
RTX 2080 - 31.8fps
RTX 2080 ti - 41.7FPS
RTX 3080 - 56.47fps (177.6%)
Doom Eternal (4k/Ultra Nightmare)
RTX 2080 - 82.3fps
RTX 2080 ti - 105.8FPS
RTX 3080 - 156fps (189.6%)
So, by my calculation (just let me know if anyone sees a discrepancy in all of this) the RTX 3080 is at most 147% faster than the 2080 ti, and at worst 133% faster. That's an incredible, and unexpected performance boost over the 2080ti that I never thought I'd see.
Borderlands 3 (4k/Bad Ass/DX12/4K/TAA)
RTX 2080 - 34.33fps
RTX 2080 ti - 48FPS
RTX 3080 - 61.79fps (180%)
Shadow of the Tomb Raider (4k/Highest/TAA)
RTX 2080 - 47fps
RTX 2080 ti - 61FPS
RTX 3080 - 79.81.59fps (168%)
Battlefield V (4k/Ultra/TAA/DXR OFf)
RTX 2080 - 67fps
RTX 2080 ti - 84FPS
RTX 3080 - 112fps (168%)
Control (4k/High/TAA/DLSS OFf)
RTX 2080 - 31.8fps
RTX 2080 ti - 41.7FPS
RTX 3080 - 56.47fps (177.6%)
Doom Eternal (4k/Ultra Nightmare)
RTX 2080 - 82.3fps
RTX 2080 ti - 105.8FPS
RTX 3080 - 156fps (189.6%)
So, by my calculation (just let me know if anyone sees a discrepancy in all of this) the RTX 3080 is at most 147% faster than the 2080 ti, and at worst 133% faster. That's an incredible, and unexpected performance boost over the 2080ti that I never thought I'd see.
Last edited:
Better compression algorithms?
That said I'd expect performance to scale 1:1 with memory overclocking. There's a reason why the 3080 isn't 3x 2080.
The lower wattage yet still so much power has me excited for the 3070. My 2k photo editing monitor will easily run that.
Like, these cards + DLSS is all I'd ever need, but this is still bullshit. It really is straight up killing SLI.
Looks like we're good with high-frame rate 10/12Bit UHD+, too!350w, 320w, 220w
According to the DF video:
Borderlands 3, 4k, maxed out:
DF 2080: 34 FPS
DF 3080: +80% = 61 FPS
My 2080Ti OC (~18 tflops): 49 FPS (+44%)
This would make a stock 3080 25% faster than a fairly well overclocked 2080Ti (in this game). That's very cool!
Control:
DF is able to get locked 60 fps at 2160p (medium DLSS), ultra.
My 2080Ti is getting ~48fps in the "corridor of doom" at those settings.
Locked 60 means a minimum of 60 and we have no idea about the average. But even that is again at least an improvement of at least 25%. Which is also nice.
Borderlands 3, 4k, maxed out:
DF 2080: 34 FPS
DF 3080: +80% = 61 FPS
My 2080Ti OC (~18 tflops): 49 FPS (+44%)
This would make a stock 3080 25% faster than a fairly well overclocked 2080Ti (in this game). That's very cool!
Control:
DF is able to get locked 60 fps at 2160p (medium DLSS), ultra.
My 2080Ti is getting ~48fps in the "corridor of doom" at those settings.
Locked 60 means a minimum of 60 and we have no idea about the average. But even that is again at least an improvement of at least 25%. Which is also nice.
Last edited:
In the reddit AMA a Nvidia rep said it was a few percentage:
Reddit - Dive into anything
www.reddit.com
So about 40% better than a 2080ti? Wow this is what I dreamed of but didn't expect. I remember some rumors indicated only a 15%-25% increase.
Thanks for posting this. I was worrying that my year old 9900k system may bottleneck this. I will probably look at upgrading to next gen intel when the next Nvidia architecture comes out in a few years.In the reddit AMA a Nvidia rep said it was a few percentage:
Reddit - Dive into anything