GeForce RTX 2080Ti feature

Official gaming benchmarks for NVIDIA GeForce RTX 2080Ti & RTX 2080 show amazing performance increase in twelve games

NVIDIA has provided reviewers with samples for its new graphics cards, the GeForce RTX 2080Ti and RTX 2080, and has included a guide which features some official gaming benchmarks. Now while we strongly suggest waiting for third-party benchies, these ones will give you an idea of their performance on modern-day titles in 4K on Ultra settings.

In Battlefield 1, the NVIDIA GeForce RTX 2080Ti is 48% faster than the GeForce GTX1080Ti and in F1 2018 the new high-end GPU is 44% faster than the GTX1080Ti. In Call of Duty: WWII we see a 52% improvement, in Rainbow Six Siege we see a 49% boost, in Mass Effect: Andromeda we see a 41% difference between them and in PlayerUnknown’s Battlegrounds we see a 36% performance boost. Simply… wow.

The NVIDIA GeForce RTX2080 is also faster than the GTX1080Ti in all of the benchmarked titles. However, and as you will see, in Shadow of the Tomb Raider the NVIDIA GeForce RTX 2080Ti was unable to offer more than 60fps on Very High settings in 4K. It came close, 59fps, but since this is the average framerate we are pretty sure that there are drops below that.

Nevertheless, the performance boost – at least according to these graphs – is tremendous between the GeForce RTX 2080Ti and the GeForce GTX1080.

Here are the benchmarks and the settings that were used for each title.

DirectX 12 Games

  • Battlefield 1 – Ultra Preset
  • Hitman – Highest Settings
  • Shadow of the Tomb Raider – Very High Preset, TAA
  • Star Wars: Battlefront II – Ultra Preset

DirectX 11 Games

  • Call of Duty: WWII – Render Resolution: Native, Pre-T2X Resolution: Native, Post Process AA: Filmic SMAA T2x, Texture Resolution: Extra, Normal Map Resolution: Extra, Specular Map Resolution: Extra, Sky Resolution: Normal, Shader Preload: On, Anisotropic Filtering: High, Shadows: On, Shadow Map Resolution: Extra, Shadow Depth: High, Screen Space Shadows: Always On, Screen Space Reflections: High, Cache Sun Shadow Maps: On, Depth of Field: High, Motion Blur: High Quality, Screen Space Ambient Occlusion: Hemo AO, Medium Distance Ambient Occlusion: On, Surface Scattering: On
  • F1 2018 – Ultra Preset, TAA, AF: 16x
  • Mass Effect: Andromeda – Ultra Preset
  • Middle-earth: Shadow of War – Ultra Preset
  • PlayerUknown’s Battlegrounds – Ultra Preset
  • Rainbow Six Siege – Ultra Preset
  • The Witcher 3 – Ultra Preset

Vulkan Games

  • Wolfenstein 2: The New Colossus – Mein leben Preset

Kudos to our reader Metal Messiah for informing and Videocardz for sharing the slides.

61 thoughts on “Official gaming benchmarks for NVIDIA GeForce RTX 2080Ti & RTX 2080 show amazing performance increase in twelve games”

  1. Please kindly check this Twitter post as well….It’s a legit source (link as given below).

    Michael @Nestledrink: On average across 14 benchmarks on 4K SDR only:

    * 2080 Ti is 45% faster vs 1080 Ti
    * 2080 is 15% faster vs 1080 Ti
    * 2080 Ti is 84% faster vs 1080
    * 2080 is 46% faster vs 1080.
    .

      1. m/ awesome!!!! 🙂 m/, thanks for the post/info man. I find it funny how so many thought the cards where a joke. I always thought it had to be close to 50% over a 1080 Ti or else it would not make any sense(they would never release them for just the ray tracing if the rest sucked).

        Sometimes gamers treat Nvidia like they are a new indie dev just starting out, I don’t know if that is funny or sad.

      2. Those are the exact numbers I mentioned the week of the RTX reveal btw… You know when everyone was losing the $hit over the Tomb Raider video…

        1. *everyone losing their sht because of RTX performance just like when Tesselation got out everyone and their mothers were ransacking the streets and now everybody uses Tesselation and there’s next to no impact. Raytrace will eventually get to that. Not this gen now. So just turn that sht off 😉

        1. Should be, but I’m not fully sure.

          Wait till these RTX cards hit the market, then hopefully the Pascal GPUs might see some decline in pricing as well, if Nvidia allows this to happen.

          As of now, the pricing is all over the place.

      1. Nope, sadly no info leaked by the reviewer..More slides should be coming soon though….

        But I don’t expect a very high TEMP of these new RTX cards, because NVIDIA themselves claim their new design is 5x quieter, and runs 20-15C cooler than the predecessor.

        But the picture won’t get clear, unless we get real TDP and temp values under full GPU load, when proper gaming benchmarks come out.

        1. Yep, it’ll be interesting to see temps given that Nvidia’s own cooling solution for the Founders Edition card pictured has evidently been beefed up.

    1. See MM you just had to have faith my man… And these are all at stock clocks! And apparently the 20xx series overclock like mad (better then 10xx series!). I will have mine watercooled for max clocks. So add about 10-15% to these scores!

      1. I agree on that.

        The overclocking headroom is pretty impressive with RTX, due to better power delivery, along with higher quality VRMs have provided more headroom for overclocking.

        Which GPU are you going to buy, btw ? The RTX 2080 Ti ? And what about the Waterblock, EK I presume ?

    2. if you look only at some numbers and you cant see the difference in graphics between latest 5 release of drivers there is no point to discuss this. Even a kid in 3rd grade can read some numbers from a chart. i gained 15 fps in all games i play with driver 399.07. Magic? Driver optimized? haha, nah, i still have eyes and a little brain.

  2. Results looks impressive compared to pascal, but 59fps in new tomb raider is indeed somewhat disappointing :P. I hope DLSS will boost performance in that game

    1. DLSS should definitely give a boost in performance, but I’m little skeptical of the actual image quality, and as whether the up-scaled image will be at par with what a native 4K screen actually offers ?

      Though, it should offer the same quality, imo.

      According to NV developer blog, DLSS is a new kind of neural network to find jagged edges, and perform high-quality anti-aliasing by determining the best color for each pixel, and also overall improve the image quality.

      And, as per Nvidia’s claim, this new DLSS feature offers the highest quality AA with fewer artifacts, than other forms of AA.

      So, if there’s a performance boost along with improved image quality, then it’s really worth enabling DLSS, if given this option.

        1. Bencrhmark is more demanding than the game . Played half an hour so far,
          max settings 1440p. I have GTX 970 G1 GAMING and while the benchmark
          say 46 fps max and 27 average on max settings 1440p, the game runs
          without stuttering or lag. Iam sure on 2080ti already runs at much more
          than 60 fps on max settings 1440p. But even on my 970 it looks amazing
          maxed 1440p. Iam wondering how much better it will look with ray
          tracing.!!1

      1. Bencrhmark is more demanding than the game . Played half an hour so far, max settings 1440p. I have GTX 970 G1 GAMING and while the benchmark say 46 fps max and 27 average on max settings 1440p, the game runs without stuttering or lag. Iam sure on 2080ti already runs at much more than 60 fps on max settings 1440p. But even on my 970 it looks amazing maxed 1440p. Iam wondering how much better it will look with ray tracing.

      2. So all this time we’ve been wondering how DLSS “improves performance” and I finally found the answer in the NVIDIA Turing whitepaper released today:

        “Whereas TAA renders at the final target resolution and then combines frames, subtracting detail, DLSS allows faster rendering at a lower input sample count, and then infers a result that at target resolution is similar quality to the TAA result, but with roughly half the shading work”

        So as you and I presumed, DLSS literally allows you to render your game at lower than your native resolution and get temporal AA quality better than your average TAA implementation. That means superior AA PLUS serious, serious performance increases.

        This tech has me SO excited….. I’m definitely not regretting my Ti preorder. I just hope that devs of some of the existing games I play can somehow work with NVIDIA on this and get it implemented, making it more of a standard.

        The problem is, the VAST VAST majority of dev’s consumers will not have the ability to use DLSS because they won’t have RTX cards with a tensor core. This means there’s little initial incentive for devs to implement it.

        On the flipside, if their consumers can double their frame rates at 4K in their games just by sending a bit of code to NVIDIA, then I can see lots of developers jumping on the bandwagon to get free optimization. So far the adoption rate is looking promising before the cards have even released!

        1. Also worth mentioning, NVIDIA detailed a DLSS 2x mode as well that renders that game at your native resolution and THEN does the Tensor core magic on those frames.

          So basically, with DLSS, you can render your game at half your native resolution with all of the performance benefits that brings, and get the quality of native.

          With DLSS 2x, you can render at your native resolution (normal performance) and get the quality of roughly twice your resolution as though you used SSAA with no performance cost.

          I REALLY hope this catches on…

          1. There are YT benchmarks already on “Source
            Syndicate” channel, and 2080ti is much faster compared to 1080ti even without DLSS. For example the witcher 3 77fps on 2080ti, and 43fps on 1080ti.

      3. DLSS looks like a great feature, but even without DLSS performance is very impressive. There are YT benchmarks already (just look for “Source Syndicate” YT channel), and 1080ti is destroyed even without DLSS. For example the witcher 3 77fps on 2080ti, and 43fps on 1080ti.

    2. Idk, the Titan V can do 60 something average FPS in SOTTR: /watch?v=Ycqu4Mv9Lv0

      The 2080Ti should be able to pull higher than that for sure. So Idk about those benchmark results. Maybe take a grain of salt until proper reviews are out.

          1. Yes, that’s correct.

            SOTR will have this new dlss feature, given this game is the latest title which will support RTX as well.

            But I think the developers haven’t implemented this feature yet, and Nvidia also needs to release a proper driver which should be having proper support for DLSS.

      1. Dont worry we will win it for free on the nvidia giveaway so iwill keep my gtx 970 until nvidia sents me the RTX 2080TI that i won for free. Just because i have the money dosent mean i must pay 1500 euros for a graphics card.

  3. So it seems the 2070 should be performing within 1-2% of the 1080ti….now im ok with the pricetag of that card. I dont plan on going 4K anytime soon(144hz/1440p is still a great sweet spot) so i wouldnt need something like the 2080ti

      1. I’ve got all these credit cards handy. I’m buying them all on credit. Only 24% interest. Then we’ll see who the real !diot is.

        1. You’re too late…. I got them all 2 minutes after the announcement.

          I stroke Donald Trump’s ego (and occasionally his hair) and in return, he funds my technical ventures.

    1. So how do the cards compare with the raytracing mode enabled on the RTX vs the GTX 1080TI? Seems like at that point the 1080TI beats both the RTX cards in framerate. Which begs the question, is the raytracing feature worth the extra money at this point?

      1. If raytracing effects will run below 60fps in new games I think most people will turn them off, but maybe it will be usefull in old games. Quake 2 with something similar to RTX looks better than quake 4 :P, and I wish to play half life 2 with RTX effects. RTX performance in battlefield 5 looks already looks solid, and they didnt optimized that game for 2080ti but for titan volta (card without RT cores). Also according to digital foundry RTX effects are rendered with 1:1 with native raster resolution, so going from 1080p to 1440p or 4K results in huge performance penalty but if developers will render RTX effects at 1080p or maybe even lower resolution performance should be good at higher resolutions as well.

        Right now the most exciting thing for me is DLSS. If DLSS picture quality will look as good as native then probably most people will use it, and with DLSS 2080ti will be 2x faster in new games tham 1080ti and run all 4K games with ease.

    2. If they were priced at 500 for the 2080 and 600 for the 2080 Ti, I would agree with you.

      Having said that, they are WAY overpriced compared to the 10 series for a pretty small gain in the 2080’s case.

      Because if I’m willing to pay 1080 Ti / 2080 prices for a GPU, then I want a little more than just 10 fps gained.

      *but muh raytracing!*

      As someone who owned a 290X, but muh TrueAudio. If it gets adopted by devs it’ll be just as great if not better on the 3000 series, and if not then I won’t have wasted my money early-adopting a technology that didn’t pan out.

  4. And as far as the Tomb Raider scores are concerned… An overclock will get the card up to an average of 67fps… So you may still see some dips but nothing to worry about. Add to that more mature drivers and it should end up safely in the 70’s…

    1. Exactly what i’m gonna do.I’ll just wait some more months,then i’m gonna replace my old good lovely Gtx 960 4Gb (still great gpu’s at that day), with an 1080ti. Time has come for my Granny 960 🙂

  5. someone start a campaign to fund me a gtx 2080ti, all you have to do is share the link and donate $5, in return 1-i won’t sell my kidney 2-i will give my gtx1080 to a poor (and proud) brown kid in middle east because middle easterners are more poor than african kids and 3-i will pray that allah would watch over you.

    1. aren’t the Suadis rich?
      or is it just like any other Islamic $hithole where the majority of the population is dirt poor?

Leave a Reply

Your email address will not be published. Required fields are marked *