Table of Contents

VR Gaming

So this is where it gets a little more tricky. Measuring performance on 2D applications is relatively straightforward, count the frames rendered in a fixed period of time using a fixed benchmarking scene, rinse and repeat. Here’s why that isn’t an option for VR.

See Also: Oculus Rift Review: Prologue to a New Reality
See Also: Oculus Rift Review: Prologue to a New Reality

1) There are no VR specific synthetic (or otherwise) benchmarking applications (with one exception) dedicated to testing your system’s performance.

See Also: First Look: VRMark Benchmarking Suite Analyzes VR Headsets & Hardware
See Also: First Look: VRMark Benchmarking Suite Analyzes VR Headsets & Hardware

2) If you try to apply 2D gaming benchmarking methodologies directly to VR you’ll come unstuck. Desktop consumer VR systems clever rendering tricks to present a consistent display update. In the case of the Oculus Rift, the headset we chose to benchmark with here, the most problematic is Asynchronous Timewarp. This is a technology which aims to smooth out any lumps in rendering performance and resulting nausea for users by drawing any given frame using the most up to date position of the user’s head, interpolating the final image to match. So, even if your GPU is falling behind on consecutive frame updates, the user may not notice as they’re always presented with a view that their brain is expecting relative to their head’s movement. What this means is that any drops in frames as seen by benchmarking metrics gathering software may not result in an experiential issue for the user once inside VR.

Asynchronous timewarp (ATW) is a technique that generates intermediate frames in situations when the game can’t maintain frame rate, helping to reduce judder. However, ATW is not a silver bullet and has limitations that developers should be aware of.

Additionally, good VR applications are (very wisely) built from the ground up to prioritise frame rate over visual fidelity, to make sure that those on the lowest recommended hardware will still receive a smooth, nausea-free experience. That focus on stability of frame rate has meant that developers have been reticent to include the pages of graphics options and settings PC users are accustomed too – the last thing you need as a developer is another set of variables to test when the results may end up in people being sick.

The upshot of all of this is that both the GTX 980 Ti and GTX 1080 are supremely overqualified in terms of raw grunt to handle the majority of VR experiences out there. But, if you’re spending almost 600 dollars on a new GPU, you’ll want to know how much headroom the card will deliver and whether you can improve your VR experience in any way to take advantage of that extra power.

So, we chose to pick a handful of VR games and experiences which allow a level of visual tweaking, to levels that will make these cards sweat and deliver a visual experience that’s a cut above the norm. Each title was run with the headset sat in a fixed location in view of the Oculus sensor, angled to catch the lions share of the action. We then gathered metrics on GPU load and the resulting frame rate to see how well each card coped. For each settings level on each card, we also ran the same test wearing the headset to gauge the actual user experience to sanity check the (potentially problematic) raw numbers.

Varjo's Enterprise Cloud XR Streaming Platform Now Supports Quest 3 and Quest Pro

Henry – Oculus Story Studios

henry-oculus-story-stdio-2Whilst the first real-time VR film experience from Oculus Story Studios may not seem like a taxing choice for this line up, in its efforts to ape the fidelity on display in pre-rendered CG movies, high levels of character detail, lighting and complex animation result in a fairly taxing task for GPUs. But Oculus Story Studios recently released details of an undocumented feature of Henry which, with one small tweak, can up to double the pixel density the experience is rendered at, further pushing the host GPU and delivering an even more ‘Pixar-esque’ experience, free from jaggies and rich in detail. This despite the fact that the consumer Oculus Rift’s viewable image could be regarded as decidedly low-rez compared with gaming desktop monitors (the Rift has a combined resolution of 2160×1200 or 1080×1200 per eye, not all of which is viewable).

First up, we ran the experience at the standard 1.0 pixel density configuration, to check the baseline performance for both cards and gauge rendering head room for both. You can see that, although neither card struggles at all, the GTX 1080 is clearly finding the task less of a chore than the GTX 980 Ti. (Click graphs for larger versions)

gtx1080-Henry-10xPixel-GPUUsage-FrameRate gtx980ti-Henry-10xPixel-GPUUsage-FrameRate

We then configured the experience to render at double the standard pixel density, resulting in a visibly sharper more vibrant image – rendering aliasing almost imperceptible. Whilst the GTX 1080 displayed a high level of GPU usage, frame rate relatively stable (if not perfect) throughout and this showed in the experiential test. The 980 Ti however ran into some trouble. See below that the card seemed to struggle to maintain a consistent experience, with frame rates taking a dive quite early on (as those detailed character models entered the scene). The results were a little confusing, so much so that I re-ran the test several times. Despite the card not obviously hitting any thermal or voltage limits, the performance took a dive every time.


In terms of the actual user experience however, it was comfortable for both cards. There was no obvious stutter exhibited, although character animation looked to become less fluid – perhaps interpolated through async timewarp, with the overall feeling of fluidity clearly impacted throughout for the GTX 980Ti. The GTX 1080 has the greater headroom outside of these issues, going by GPU usage alone and felt completely fluid throughout.

Project Cars – Slightly Mad Studios

project-cars-e32014-5Project Cars represents a graphical workout in its vanilla gaming form and that translates to VR. The game is one of the few early VR titles which allows the user to customise at the same levels as a standard PC game. The VR defaults however are pretty ugly, with heavy aliasing evident and not much in the way of texture detail or indeed lighting fidelity. Again, in the interests of testing the VR rendering mettle of these cards, we maxed out almost all ‘Performance’ tab settings available, and only disabled or dialled back the distracting ‘Visual FX’ such as Lens Flare. We then tested 3 passes on each GPU with successively greater levels of anti-aliasing, from No AA to DS4X right up to DS9X – each clearly more taxing than the last.

Deadmau5 Concert Experience Comes to Social VR Music Venue 'Soundscape'

With the Oculus Rift headset positioned in a marked position on a desk in sight of the Oculus sensor, angled such that the view captured the majority of the road ahead. We then ran a fixed, one lap Solo race, with 19 competitors at Oulton Park ‘Fosters’, using the monitors mirror view to steer the car.

First up, we ran with no anti-aliasing. Even with most graphical options at their highest, in VR the resulting aliasing was extremely distracting using these settings. Nevertheless, Project Cars represented a significant VR challenge with both the GTX 1080 and 980Ti pushing high GPU usage levels. Despite the large troughs in performance for both cards (mainly when all cars are bunched up in game), stuttering was not an issue, fluidity was affected until the 2nd half of the lap, with slight ghosting on opponent cars. All in all, the experience on both cars was far better than the below graph might lead you to believe.


Next, the same run with DS4X AA engaged and a much improved image. With jaggies reduced (although still evident) it’s easier to focus on the environment although the horizon was still ultimately a blocky mess. As for in-headset performance, the 1080 provided the more fluid experience although neither card was able to maintain a rock solid 90FPS. Surprisingly, considering the variation of frame rates, there was no stutter evident, but movement appeared far less fluid on the 980Ti with a distracting subtle ghosting effect around moving objects.


By the time we reach maximum AA settings at DS9X, the 980Ti is struggling to maintain half of the required 90FPS frame rate. The GTX 1080 is also showing signs of strain but the resulting gameplay inside the VR headset is now refreshingly free of aliasing and in fact could be startlingly immersive. Again, Oculus are sprinkling their magic pixie dust, with the low frame rates on both cards still resulting in a playable, if less than ideally fluid experience whilst in the Rift. For those with strong VR legs however, this high fidelity rendering option does provide a striking, presence inducing image on occasion.


Chronos – Gunfire Games

chronosChronos, with it’s static camera based viewpoint is another title you wouldn’t necessarily associate with pushing performance boundaries, but it’s a beautiful game and its opening cinematic is reasonably taxing – especially when you push the in game settings to ‘Epic’ across the board, that includes an increase of overall pixel density. With all that said, Chronos is clearly a very well optimised game. Even in this maxed-out configuration, neither card faltered rendering the opening sequence. As with previous benchmarks, each run was performed with the Rift positioned on a desk in view of the Oculus sensor, angled to capture as much important detail in the scene as possible.

Both the GTX 1080 and 980 Ti provided a stutter free user experience overall and that’s born out by the rock solid 90+ FPS figures seen below.

'Batman: Arkham Shadow' Trailer Reveals Story Behind Quest 3's Next Big First-party Exclusive


However, looking at the GPU usage gives you an insight into just how much more headroom the GTX 1080 has compared to its older brother. It’s not an insignificant difference, with the 1080 barely passing 60% and the 980 Ti clearly sailing north of 80% during the more demanding moments of the sequence, most notably when dynamic lights were at maximum.



The GTX 1080 is here and it’s the most powerful single card Nvidia have produced yet. Raw rendering performance gains over the established Maxwell-based GTX 980Ti, whilst not epoch making, are significant – especially in our DirectX 12 benchmarks. In 2D applications the card can push high resolution, super-wide gaming admirably and when it comes to VR gaming, gives you some high quality options that are perhaps just out of reach of the 980Ti’s capabilities.

All of this is at stock frequencies too. As this new Pascal 16nm GPU is extremely efficient (on load temperatures of around 55-65 degrees centigrade or less) mean this card will likely provide a tasty target for overclockers. A test version of EVGA’s Precision X tuning tool turned up too late to include in this article, but the card was easily able to sustain a 200MHz base overclock with some voltage tweaks and I suspect it’s capable of much more.

As for those claimed fantastical VR performance gains, they have to be treated as largely theoretical for now. Pascal does introduce a radical, hardware assisted technology in the form of Simultaneous Multi-Projection, a parent technology which powers many VR relevant functions such as Multi-Res shading and Lens Matched Shading and all form part of the recently re-branded VRWorks suite of APIs. We’ll dive into more detail on some of these technologies imminently. But the point we need to make in the mean time is that, whilst all of the above could bring huge performance wins for virtual reality applications, they all require software to be coded to support them and, by definition are Pascal and Nvidia specific. And whilst Nvidia are working on wooing VR developers to adopt VRWorks right now, those performance benefits may take a little while to surface in the real world.

Owners of GTX 980’s will see a substantial uplift in performance should they choose to upgrade and those 980Ti owners with deep pockets who are willing to dabble in overclocking will also see very worthwhile benefits too.


If your focus is purely VR however, the upgrade argument is perhaps less compelling at this stage. With application developers judiciously targeting performance well below the capabilities of the GTX 1080 for very good reasons, it’s hard to see the need for 1080 levels of raw grunt for a little while. However, if you do upgrade, you will find yourself with the performance headroom to maximise visual fidelity in VR for some titles right now and potentially more down the line.

Disclosure: Nvidia paid for travel and accommodation for one Road to VR correspondent to attend an event where information for this article was gathered. Nvidia also provided Road to VR with a GTX 1080 GPU for review.

Newsletter graphic

This article may contain affiliate links. If you click an affiliate link and buy a product we may receive a small commission which helps support the publication. More information.

Based in the UK, Paul has been immersed in interactive entertainment for the best part of 27 years and has followed advances in gaming with a passionate fervour. His obsession with graphical fidelity over the years has had him branded a ‘graphics whore’ (which he views as the highest compliment) more than once and he holds a particular candle for the dream of the ultimate immersive gaming experience. Having followed and been disappointed by the original VR explosion of the 90s, he then founded to follow the new and exciting prospect of the rebirth of VR in products like the Oculus Rift. Paul joined forces with Ben to help build the new Road to VR in preparation for what he sees as VR’s coming of age over the next few years.
  • linagee

    “The GTX 1080’s base clock of 1607Hz boosts at stock to 1733Hz under load”

    Surely MHz instead of Hz?

    • RoadToVR

      Heh, indeed. Altered. Thanks!

  • Badelhas

    Great article!
    So, for VR only, you dont believe its worth spending 600 dollars on this
    card and we should go for a cheaper 980 instead?


    • Mexor

      The GTX 1080’s architectural advantages in VR over the GTX 980 need to be specifically programmed for by developers. Obviously that hasn’t happened yet, and how much it will happen in the future and how soon is unknown, I suppose.

      • Rogue_Transfer

        Any future games(or existing ones using Vorpx) that are designed with more detail will utilise this. In particular, this will be vital for planned big AAA games that incorporate a VR option – like Star Citizen, etc. to present with their normally intensive graphics that current GPUs without Pascal may not manage in VR and would require more work to downgrade assets/effects for.

        • Mexor

          Well I hope so. But I wonder how much work developers will be willing to put into features that, in the near future, will be addressing a comparatively small market. The GTX 970 has been enormously popular and is VR capable.

          I also haven’t seen any analysis of the performance boost versus image quality trade-off that SMP for VR has, or even if there is any trade-off. I myself have no idea how accurate the current algorithms used to invert the lens aberrations are, but theoretically it should be possible to achieve a level of correction unattainable by the 4 linear approximations (per eye) offered by SMP. But maybe those approximations are simply good enough. Perhaps someone in this forum has a better idea and can clue us in.

          • Rogue_Transfer

            Unfortunately, no analysis can be done of SMP until GTX 1080 and GearWorksVR are released and integrated into existing VR APIs like SteamVR/OculusVR/Vorpx.

            The approximations being sufficient is an important point, it’ll be interesting to see how well they actually work in (virtual) reality. Plus, since SMP can do up to 16/32, it makes me wonder if more than 4 could be used (per eye) to offer a better curved approximation or is this number integral to the graphics driver for VR use.

            Remember, this Pascal feature is likely for the whole range of GPUs coming out this year and next and with the improved performance all round, even lower bracket Pascal cards may usurp the GTX970 as recommended min. spec. over that period. Esp. if other manufacturers release improved headsets (this year or early next year) before the refresh of the Vive & Rift occurs.

            Minimum specs vary per game and although the guidelines from Oculus/Vive are for a min. 970 for the lifetime of their VR headsets, that’s not going to stop developers offering an option to turn on greater detail that normally is for lower frame rates on monitors. It’s the triple-A games coming with both monitor & VR support that will utilise this the most, because without it, the older generation cards can’t cope with the desired detail at 90 FPS.

            Existing, lower performing cards without this feature just may either not be supported in VR for these titles to come, or run with a much reduced quality level to keep the frame rate up or run a lower native resolution and upscale to the headset. VR already is a tiny market (due to the HW costs involved), mostly enthusiasts with money to spare and they are (generally) more than willing to upgrade GPUs quicker than most, if there’s a feature that offers ~2 times the performance in VR, as SMP is claiming.

            The VR headset is on a fast track for upgrades (between mobiles yearly refresh and consoles). These will require greater performing GPUs and the old generation ones won’t cope soon enough. Whether AMDs next gen cards will offer similar benefits for VR we’ll find out soon, but if not and if SMP is anywhere near as good as claimed, nVidia Pascal cards will be the logical VR choice for that desperately needed potential for 2 times the performance to support detailed scenes & effects over simplified ones.

            For developers, doing VR without this feature & potential performance will be much more work. They already know they’re targeting a new and fledgling market that’ll likely grow based on game min spec. demands, rather than the ones set by the VR headsets. Esp. considering the new Pascal cards are replacing higher cards for the same price. Most of the potential VR market purchasers hasn’t been tapped and just now are choosing their GPU hardware.

    • Jason Evangelho

      I’d actually suggest waiting for the GTX 1070.

    • Joel Barsotti

      If you have the money to spend, get the 1080 for sure. If you want to spend less, I’d hold out for the 1070, it has the same architecture changes to support VR should be 980ti performance in VR.

      • RoadToVR

        Or, in fact, what he said. :)

    • RoadToVR

      The two choices you have if you’re upgrading from an older card are whether to wait for the cheaper GTX 1070 just around the corner in June or spend more for the 1080 based on the scores you’ve seen here and elsewhere. I wouldn’t buy into a 980 now – Pascal offers too many future benefits.

      • Badelhas

        Thanks for your answer! That’s what I’ve been thinking. I probably know what you are going to answer, but do you believe that buying the Vive is worth it? ;) The lack of proper complete software (games) and the screen door effect worries me. I don’t want it to happen like when I invested in a projector, high end pc and NVIDIA 3D vision kit. I thought that would bring back my taste for pc gaming but it had a short term effect…

  • Ellathar NithranLunelor

    Crazy GPU enthusiast says thanks for that well written article.

    • RoadToVR

      Many thanks! Glad you liked it.

  • Sam Illingworth

    Hmm, do I then take it that Project Cars isn’t going to be playable on my 980 (non-Ti) in VR?

    • tineras

      This person ( is playing on a 970. So you can play, but may need to bump the settings down. I have a 980 as well (2, but no SLI for VR), so I am going to be upgrading to a 1080.

      • Sam Illingworth

        Well that looks good to me. Is 60fps alright in VR then, and are the visuals in that 2D video representative of what we’ll get in the Vive? the article suggests otherwise.

        • tineras

          I have a Vive and Project Cars, but I haven’t actually tried it out yet. I ordered a sim seat that I’m expecting to arrive on Friday. I’ll try it out then.

          My expectations for visuals are that they will look similar to the above video at somewhere between 480p and 720p. But that’s just my guess based on other games I’ve played on the Vive.

          As for framerate, I have no idea how well the Vive’s reprojection will compare to ATW on the Rift. *fingers crossed* that it’s a good experience. All the feedback on the Vive subreddit about PC is positive, so that’s a good sign.

          • mbze430

            Are you running SLI?

          • tineras

            For project cars, you have to disable SLI or everything is screwed up. And for VR in general, SLI is not supported. So while I do have 2 980s, I do not run SLI. I plan on buying a 1080 soon to replace them.

        • tineras

          Hmm… it turns out that performance might actually be a huge issue. Have a look at some of the comments here ( The person in the vid had their settings all the way down. And users of 980ti and 980 are reporting poor performance as well. Bums me out…

          • Sam Illingworth

            Hmm. I’d like to try a demo before I buy it I think…

        • eck

          If your computer can’t draw 90 full frames every second, it’ll drop down to 45 per second (with the reprojection). There’s no 60, at least in Project Cars.

          • MasterElwood

            Only on the VIVE – because their timewarp sucks compared to the RIFT´S ATW.

      • mbze430

        That looks like a lot of setting being turned down. I have a 980TI setup with everything on MAX/ULTRA/TO THE LIMIT, except for DSxA (just regular MSAA) at 4k and I can do 60fps+

        But I ALWAYS have been wondering how it will do in a 3D/VR with the Oculus (Waiting on mine). I think it will be fine with DS9A… but now I worry about latency…..

        Anyone out there with a VR setup running SLI/Crossfire??? I need some input

    • SeanC

      It’s playable with a 970

      • Sam Illingworth

        Playable as in a VR-suitable framerate with decent graphics? the above seemed to suggest even these cards failed to run it well with enough AA for it not to look horrible. Unless I’ve misunderstood?

        • Joel Barsotti

          They were playing with maxed out settings in the benchmarks above.

          Lowering things like shadows, car detail, track detail, ect… can have a big impact on frame rate. I played it with a DK2 back before async timewarp and it was good but not perfect then. I think with async timewarp and the right settings the 970 should work well.

          • Sch@dows

            Indeed, I play it with a 970 (and a vive) and it is perfectly payable (thank you retroprojection / atw).

            Obviously, my setting are a mostly on LOW with a few bit of Medium here and there, but frankly, except for the AA (stuck at DS2x), most the settings don’t have that much of an impact visually due to the poor helmet resolution.

            You will have to choose between greatly improved immersion, or more realistic graphics and full grids.

    • RoadToVR

      As stated in the article, despite some pretty significant drops in framerate, it was still playable with some at times jarring artefacts. As I say, depending on the strength of your VR legs, the extra visual punch may be worth the drop in frame rate.

    • tineras

      I just wanted to follow up and let you know that I tried PC on my 980. The graphics settings are quite low by default but it honestly didn’t bother me. After a few minutes of racing, I was focused on the fun of the racing and not the resolution/graphics settings. I tried a practice race with no other cars on track and the performance was perfect. I then tried a few races with 12 AI opponents. I noticed a little bit of jitter when turning my head with other traffic around, but nothing that actually made it unplayable. Cars tend to spread out after the race starts anyway.

      All that being said, Assetto Corsa is now supported in the Rift (and Vive via ReVive). Devs said that native support for Vive is coming, but didn’t provide a date. Users are reporting that performance is excellent even with high settings. I can’t wait to give that one a go as well.

      One more note: SLI is not supported by Project Cars or any(?) other VR games for that matter. But it’s really screwed up in PC, so I would disable it first. I forgot to and was quickly reminded.

      • Sam Illingworth

        Thanks for the info. I think I’ll put this and AC on my wishlist, and maybe pick one up when they’re in the sale, when native Vive support for AC comes out or when PC gets patched to make it run more smoothly. There are quite a few games I’m looking forward to on the Vive, I don’t need to rush into anything :)

      • Sam Illingworth

        AC 40% for the next eight hours! I’ve bought it, let’s hope Vive support isn’t far off.

        • tineras

          Can’t remember if it was already mentioned, but you can play now using ReVive ( I have AC, but haven’t tried it yet.

          • Sam Illingworth

            Thanks, I might have to give that a try! I was hoping not to have to install Oculus software as I want to keep my PC as clean as possible, but if it gets this working maybe I can’t wait! Could also play some if true Rift exclusives, but I’m reluctant to give them money for those…

  • Pistol Pete

    So the 2x VR performance of a Titan X for Pascal was indeed all smoke and mirrors bullshit!

    • Slam

      It’s a feature that has to be supported by the developer, which will be turned on in the Geforce Control panel

      I would imagine support for the feature will arrive soon. But it is dependent on the developer utilising it.

      • John Miller

        Wait 6 mths make sure games are using it

    • VRRV

      I believe the 2X VR performance was compared to 980sli, (they simply said better than the titan) in their presentation. Which is still misleading since hardly any VR apps have incorporated sli optimization. I read in another review today that 980sli vr performance is actually worse than a single sli because it creates more latency without reaping any of the sli benefits since no one is developing vr for it.

      • Sch@dows

        The “2x perf and 3x efficiency” was against the Titan X (see 1st page of the article).
        But that was using SMP and LMS, which isn’t supported by any games for the moment.

      • mbze430

        are you serious? that is not cool. I have a 980SLI setup (waiting for my Oculus to arrive) to only mainly play Project CARs. How bad is the latency????

        • Jona Adams

          I have to disable SLI to play PC in VR. Otherwise all the cars look jittery.

          • Bruce

            Ive played project cars on my cv1, and Im using 2x 770s in sli. Have you updated your drivers?

          • Toothlover

            Really? how many frames do you get?

          • Bruce

            OK I was wrong! I’ve got 2x 770 gtx 4gb cards, and i’m only using one! The second is doing physx or somthing. And it’s smooth in assetto corsa and project cars. I’m quite shocked actually I thought it was using both! I guess one is doing some physics, but it does kind of show that the 970 minimum recommendation is overblown.
            The cards are superclocked ones. But still…
            When I updated the drivers, it disabled sli, and I tried assetto corsa and project cars thought wow! I’m not going to need a gtx 1080 after all!

          • Toothlover

            I have 2 680s… I wonder if that would apply to me too :P

          • Bruce

            Definately worth trying it before going mad and buying one of these new expensive Nvidia cards. You might not need to upgrade.

          • Jona Adams

            At the time that I was having issues, yes. I’ve been building game rigs since 1998. I got the basics down.

          • Toothlover

            What are your specs?

          • Jona Adams

            i7-6700k, 12 Gb RAM and when I had 2 cards, 980’s in SLI.

            It’s not a system thing, it’s just a fact that no VR game at this present time supports SLI.

            Which is why I jumped on the GTX 1080 bandwagon. And dropped about 200 watts of power usage.

            I’m into high performance, low power, zero noise builds.

        • Taylor Stoll

          It’s not latency… SLI causes a sever “ghosting” of the image. I run dual GTX980’s and I can attest that SLI does not work. Again it has to be activated via the devs. using VR Works. Instead of AFR the cards each produce an image for each screen (eye). That is why SMP allows a 2x increase in performance. Instead of the card needing to produce an image for each screen individually SMP allows the Pascal cards to do the same (separate images for each screen) in one pass instead of two.

    • I think they have limited it so that it is marginally better…

    • Taylor Stoll

      No it is NOT BS. The games that will allow the GTX1080 to have that 2x advantage are game that are made with Nvidia VR SMP technology. The games only need to be patched for it to work (they may need to be built from the ground up but I have been reading that may not be necessary). Plus I think something is VERY wrong with some of these benchmarks. pCars on the 1080 should be running far higher then the 980ti. Something is wonky there…. Maybe it’s just immature drivers. But there is no way that the 980ti should be performing better then the 180 at any point during that graph. Odd….

      • Bruce

        Because overclocked 980ti are still very good. And as the multiprojection is gameworks vr, and now that amd has the console market sown up, it does make me wonder how many games will be using gameworks now.

    • oplix


  • Pistol Pete

    Why no testing with the HTC Vive RoadtoVR, it obviously the more popular VR set???

    And I have both HMDs btw.

    • RoadToVR

      Mainly time constraints. We only got the card last week and this stuff is super time consuming. Also, the games highlighted lent themselves well to being tweaked and benchmarked.

      Ultimately it’s immaterial, it was about testing the GPU not the headset – any performance gains seen here would be transferable to the Vive.

      • crumb

        Wouldn’t the steamVR vs oculus home make a difference though? If you do get a chance, I’d love to see benchmarks in The Lab or the Galary.

      • Pistol Pete

        That’s for the response RoadtoVR. I do understand these benchmarks do take a long time and I do greatly appreciate you taking the time to complete them.

      • Matt R

        That may not be true at all. The tracking systems use a very different amount of processing power. It would be interesting to see if the very efficient way the Vive calculates the tracking has a positive effect on frame rate compared to the very resource heavy image based tracking for the Rift. I also own both HMD’s

        • RoadToVR

          Whilst that is true to an extent, the overhead is as I understand it negligible and besides, as I say, we wanted to test GPU performance not general system performance. It’ll be interesting to see side by side tests when real VR benchmarks become available though to test the tracking processing overhead in more detail.

    • MasterElwood

      Only about 50.000 Vive´s out there.

      • phaethon

        They are selling like hot cakes:)

  • Thomas Phifer

    Thank you for this article. I was kicking myself for buying the 980Ti right before Pascal dropped but now I don’t feel so bad. It does, however, look as though the value is greater for the 1080 dollar-to-performance.

  • JackG

    This is informative and extremely well-written – what a pleasure. I have a new i7 6700K mini-itx build sitting here just waiting on a GPU (and an HTC Vive) and it looks like the 1080 will be the way to go for me. I’ll wait on the AIB’s, though, so see how far they can push it.

  • John Miller

    Really wish the standard had been a 980 for VR

    • Bryan Ischo

      At this point, we need MORE experiences, not PRETTIER experiences, so I am OK with choosing a less expensive base requirement, because it makes it easier/cheaper for more people to get into VR, which should help translate into motivation for developers to create MORE experiences.

      I am most excited for the 1070, which supposedly is much cheaper than a 970 but still meets minimum VR specs.

  • RoTaToR

    Nice Review. As a new owner of a 4K 21:9 Monitor i will update from 780TI -> 1080.

    • phaethon

      An sli will bring you higher benefit for less price

      • RoadToVR

        Leaving aside the fact that you’ll forgo the advantages that Pascal architecture brings, and that you’ve spent the extra money on an SLI motherboard, running SLI is not analogous to running a single card. Not only do you need the game to ship with support, you also need profiles within the driver and even then you may suffer artefacts associated with Multi-GPU setups.

        • NickCoombe

          True to a degree, but don’t rule out two 1070s in SLI! I bought two 970s specifically for VR, rather than a single 980 and I’ve rarely had problems running games in SLI, and the performance gains have been very noticeable.

          As an example, I can run the new DOOM at 50-60fps on a single 970 at high settings. DOOM is not optimized for SLI (they have a “real” SLI update in the works, along with Vulcan support), but when I forced alternate frame rendering with SLI, I can now run the game at 100fps on my G-SYNC monitor at high settings with 8X AA (DOOM uses an efficient temporal AA which works really well).

          With SLI-optimized VR drivers becoming the norm, and VR being so well suited to SLI, and with the performance of the 1070 behind it, the SLI option is well worth considering.

          • phaethon

            I got a 295×2 Amd (2x290X sli-ed – crossfire) While I do not see so far any gain from the VR games, this is definitely going to change in the next couple of months. In any case all the AAA games that I am interested in support CF/SLI.

            Regarding reason to upgrade to a 1080GTX:
            Going quite arbitrary here. While a single 1080 gtx would provide a steady hypothetical 70fps (~200Watt) , a 295×2 can provide 60-62fps (510Watt). But instead of spending(~investing~) 800E to a 1080Gtx I rather spend 200E for a second hand 290X (300Watt) that would raise considerably the fps of supported games to 80fps. So taken you have the right PSU and the money for the electricity bill, this is the perfect time for an upgrade :) Especially considering what A.Vlachos presented at GDC 2016 regarding time to render a frame in a single vs multi GPU config.

            What you hear about Pascal and premptive rendering for VR to my ears sounds that NVIDIA once again has done its marketing ploy to ditch their last year’s hardware and provide software and firmware only for the latest generation so they can sell more.

            I had a GTX 690 in the past and as soon as the 700 series came out , driverwise it became irrelevant even if it was more than on par with the 700 series.

            Bad marketing tactics hurting fan base= Driving buyers to the competitor.

            In fact the 295×2 holds more to the test of time and the reason is that AMD still supports their R9 2xx series cards 3-4 years after they appeared on the market. Can’t say the same for NVIDIA…

    • oHOWEo

      You didn’t get a 4k 21.9 I’m afraid

  • Nick

    Page 3 Chronos Rift framerate, The graph series labels incorrectly shows that both lines are for the GTX1080.

    • RoadToVR

      Thanks. As soon as I can generate a new graph I’ll fix this. Luckily its on a graph that’s immaterial as to which is which as both are rock solid.

  • Sam Kennedy

    I know you guys address it but it really isnt 1:1 when you have a board partner overclocked 980 ti against a stock 1080 board partners are going to be releasing 1080s at over 2ghz which has shown up to a 25% improvement over stock so I think the real comparisons will either rely on you guys overclocking the card (although you probably wont reach the same heights as 3rd party coolers (liquid or air) )

    • Sch@dows

      as said in the testing methodology :
      “As the EVGA 980 Ti supplied in the Exemplar comes with a mild overclock, the card was reduced by 90GHz base clock to better represent a stock 980 Ti configuration, only fair as the GTX 1080 is also running at stock speeds.”

  • Mark Seymour

    Great article. This has helped answer some questions about my next PC build.

  • Sch@dows

    Seeing how the 1080 is on par with the 980Ti without AA in project CARS, the card seems limited by the CPU here. And as soon as you increasing the load on the GPU, the difference shows up.

    To see the true potential of the card, a test with a stronger CPU (8 cores 5960k for example) would have been better, but it’s true most of the potential buyers won’t have anything beyond 6600k/6700k, so the results are probably closer to what people will see.

    • PC Master

      The 5960k wouldn’t have made any difference. For pure gaming the 6700k is currently the fastest CPU on the market, and the best candidate for cards to be tested on.

      • Sch@dows

        While I would agree on most other games, that’s the most logical explanation for having the 1080 performing at the same level as the 980Ti on that test alone.

        Project CARS, as most racing simulation, is heavy on the cpu (ai, complex physics and tires engine), and the game is already known to generate an enormous amount of drawcalls (first reason for underperforming on AMD cards) which increase the load even further.

        My 4690k is already under heavy load when combined with a 970 (gpu bound scenario), remove the gpu boundary and add to that the cpu usage reserved for VR, and there is a strong possibility that particular scenario was cpu bound, hence my suggestion.

        But as I said, anything more than an Intel 4 cores cpu wouldn’t represent a typical gaming configuration.

      • speach

        Tell that to my 5.6GHz OC’d 3930K…

        • PC Master

          Oh really now? Post your CPU-Z validation, along with CPU temperatures under load, voltage, benchmarks which includes games, Cinebench, 3D mark physics scores screenshots that confirms all of this and the cooler you’re using.

          Along with a picture of you standing next to your PC holding a sign that says 5.6ghz OC 3930k then maybe I’ll believe you.

          • speach

            I don’t particularly care about Internet validation so you can take this however you want it. Here is an Arx screenshot taken during a 50 pass extreme stress mode Intel burn test on test 35:

            System is a Gigabyte GA-X79-UD7, 16GB of DDR3 @ 2400, GTX 770 4GB and as previously mentioned, 3930K. Cooler is an H100 AIO.

            After passing the IBT the system ran Prime95 blend mode for 36 straight hours with no failures across all 12 threads.

    • mbze430

      I think it’s on par because of the memory. Once AA is turned on a large amount of memory is required. Think the 8GB trumped the 6GB on the 980TI

      • Sch@dows

        Possibly, but if that’s the case, that would also means that as long as you’re not hitting the memory limit, the performances of the 1080 are the same than the 980Ti, which is inconsistent with other games/benchmarks results.
        Generally, the reason for 2 GPU with such known performances differences to exhibit the same results would be because their is a bottleneck before the GPU.

  • jacobpederson

    Is there a way to test MTP latency on the 1080 vs the 980ti? Thanks!

  • Awesome compilation. Also the performance part for the GTX 1080 is written well but if you want to compare GeForce GTX 1080 from the last versions then you should have a look here :

    Thank you!


  • Taylor Layton

    I got a GTX 980ti for 450 dollars on eBay. Do you think I should have waited and got the 1070 or is this a good deal? I didn’t want to wait but I will return if you think I should. I figured that the 1070 is probably going to be 400-410 for the SC version of a 1070 and the 980ti I bought is the SC version ACX 2.0+ for 450 dollars. So only 40 dollars more or less. Do you think 1070 will be better then 980ti single card non-VR?

    • Canaan

      GTX 1070 at stock is up to 9% faster than a Titan X at 55% less the cost.(Gamespot)

      I wouldn’t be upset about the 980TI because they will drop in price and in 6mo. to a year you could go SLI and smoke that single 1070. Always a way to work around technology these days, its kinda cool how we can throw two old ones together to come back up on top. I canceled my 1080 SC ACX 3.0 card and just bought a 2nd 280x for crossfire. I’ll wait for Polaris and Pascal comparisons first.
      Not a fan boy of either.

      So yes, I would have waited but you didn’t make a bad purchase.

  • mbze430

    Actually I have been trying to find a review on VR w/Project CARs. I am STILL waiting for my Oculus Rift (Late June shipping date now). Been wondering if my 980TI SLI setup will do the required 90fps with full AA. This review gave me a pretty good idea but you think a 980TI SLI will be able to do the full 90fps with full AA and how much of a latency issue?

    Currently I play Project CARs at 4k res, with everything max except AA at MSAA. I get over 60fps.

    As for the 1080… I think I will waiting for the “1080TI” version (SLI of course!)

    My Firestrike score and system specs:

    • Sch@dows

      Pcars doesn’t work well with SLI in VR (brings more stutters than performances increase). So, if I had to answer right now, I would say : no, your SLI setup won’t let you achieve 90fps with everything at their highest settings, including AA.
      You can still achieve 90fps with a single card by decreasing some settings with negligible impact on visuals (especially considering the helmets resolution) like Shadows, track details and crowds.

      • mbze430

        that is HORRIBLE news. I got the 2x 980TI wanting to have MAX settings for both 4k and VR. Even the 1080 isn’t powerful enough to turn on MAX…

        • Sch@dows

          Unfortunately, that’s to be expected from current games (compared to all the casual/technical demos). Maybe you will get better results with less demanding games like Assetto Corsa or iRacing.

          But frankly, what’s the point in running everything at max if you can’t tell the difference.
          Shadows are one of the most gluttonous settings, and you barely see any differences between medium and ultra settings on a monitor so I don’t even think you can see it in a VR helmet.
          Track detail mostly affect the distance at which track-side objects switch level of detail (LOD) which, again, has barely any impact on what is seen in a vr helmet considering the low resolution at which the image is rendered.
          You can leave the Car detail at ultra because the only difference with High is the level of detail of the player’s car in external view … which obviously isn’t used in VR).
          But I agree that removing the crowds is noticeable (I disable it because I don’t care about the spectators when I’m actually racing and because they generate lots of unecessary drawcalls).

          But even if you’re not maintaining a stable 90fps, ATW or retroprojection (I have the Vive) makes it perfectly playable (for me at least)

        • Buddydudeguy

          everything maxed is pretty arbitrary. There are quality settings that are simply not worth it.

  • Martin Mueller

    The NVidia Help desk told me, that I am able to activate the Stereoscopic view on the 1080 before I start any game so I could play GTA or Skyrim on HMD without using vorpx or vireio.
    Which would be nice because HTC Vive is not supported yet by either.

    In addition vorpx and vireio do have problems to display text inside the game, like Mass Effect. In the NVidia 1080 presentation was mentioned that 4 view ports per eye are used for each VR Lense, so the steroscopic view (if possible to use) could solve this.

    Can anyone confirm this?

    Thank you in advance.

    • Nathan Casey

      for every game or select? this would be amazing

      • Martin Mueller

        I have the Cybermind hires800 HMD and it was working like that, so it would be nice if someone with a vive could confirm that statement from NVidia, than no vorpx is needed with the 1080

  • wargamer1969

    No 4k benches is pffft. Plus I’m waiting on the non reference cards with better cooling and 2 x 8 pin connectors for much higher stable overclocks than the founders/ref 1080.

  • DAB

    Just having to only go through the geometry engine once instead of twice is alone going to lead to a monstrous improvement.

  • Chugs 1984

    That’s amazing. Basically that GTA V 6880 X 2880 test means that the card is pumping out 14.2 gigabits a second! Basically 6880 X 2880 X 30 fps X 24 bit colour depth = 14,266,368,00 bits per second.

    I was thinking of getting the GTX 980 Ti in order to do VR but i’m having second thoughts especially seeing the HTC Vive requires an output of about 5.5 gbps. Its only a matter of time when UHD VR is launched or they work out a way of upper the refresh to 120 fps.

    Its then you’ll need a card that can handle that

  • masterofevil22

    Does anybody know if an OC’d 980ti will beat or be beaten by an OC’d 1070? Could someone share a link please? Thx!

    Also, check out XBnPC on YouTube..if you’re into that sort of thing ;D

  • Badelhas

    Very nice article, and the first one to try to benchmark VR performance on this gpu. Not even AnandTech and Tom’s Hardware were able to do it. Congrats!
    Can we expect a similar treatment to the Nvidia 1070 and the new AMD Polaris GPU’S?


  • Gabe Trammell

    Unless you’re gaming in 4k there’s not that much of a performance gain to justify a upgrade from a 980ti to a 1080. I myself am waiting for stacked HBM ram before I upgrade again my gigabyte waterforce 980 ti works just fine everything I throw at it plays at ultra settings. I have GPU overclocked to 1480Mhz and the ram overclocked to 7800Mhz. On a 1080p set up that’s all I need. As far as VR it’s still a joke needs to be a lot cheaper for a VR headset and way less bulky. Maybe when VR is 200 to 300 dollars and way smaller I will think about it but for now it’s just a gimmick.

  • Frank Atkisson

    I’ll be sticking with my 980ti for a while yet, but I’m definitely impressed by the first run of credible benchmarks showing off the 1080’s power, as well as the pricepoint.

    My 980ti will be good enough for quite a while yet, but its nice to know that when it comes time to upgrade again, we’re still seeing comfortably significant gains from generation to generation to make it worth staying interested in doing.

  • Toothlover

    Wow, you just compared a reference GPU with a non-reference one… Just wow!!

  • oplix

    There is no benchmark software for SMP right now i believe

  • Navn_Her

    The percentages are faulty. All of them.
    Example: “Rise of Tomb Raider : 68 FPS vs 40 FPS is 50 % gain.”
    Well, if you start with 40 and add 50% to that (40*1,5 = 60), you will end up at around 60 FPS. If you start with 68 FPS, and see how much bigger that is compared to 40, you will end up with around 70 %.
    Because 68/40 = 1.7

    I hope you are not getting a pay rise in % :-P

  • Originaru

    Very interesting graphs on Henry with 2.0x pixel density, looks like the 980ti is having some problems, as 1080 not, could that be related to Pascal “async” capabilities?
    At first it seems unrelated, possibly because with 1.0x neither have problems rendering, but as it puts the cards under real stress maybe those “context switches” couldn’t be overcome by brute force of the maxwell arch.
    Just a guess, no bashing please.

  • Blasted

    I have a 980ti, it’s good enough and the 1080 is probably only 5% better; not worth the buy, especially when their DX12 support is floundering, almost as bad as Maxwell cards, no thanks NBadia. AMD this next go round. If Nvidia pulls their heads from their $$greedy asses$$, I may look at their stuff again. Until then, Bon voyage!

  • Carby

    Asus X34 predator hey lol

  • oplix

    i would love to see nvidia have an upgrade program to break the monotony of SLI