cancel
Showing results for 
Search instead for 
Did you mean: 

Quest 2 versus Rift CV1 performance benchmarks for PCVR

nalex66
Volunteer Moderator
Volunteer Moderator

I've seen people claiming lately that using a Quest 2 for PCVR comes with a massive performance cost compared to PC-only headsets. Some posted benchmarks in the Index thread implied that the Index ran 30% faster than the Quest 2 when using similar settings. This didn't seem comparable to my own experience, so I thought I would do my own benchmarks to see how my Quest 2 compares to my Rift CV1.

 

I tested both headsets using OpenVR Benchmark through SteamVR. I did my best to match pixel count between the two headsets, and ran the test at low res (Rift 100% in SteamVR), and high res (Quest 100% in SteamVR, using 1.7x resolution in the Oculus app). I tested the Quest with AirLink (dynamic bitrate up to 200 Mbps), tethered Link at default bitrate, and also with bitrate raised to 350 Mbps (the point at which I don't notice any compression artifacts).

 

EDIT: I ran the tests with Virtual Desktop, and those results exceeded both Link and Rift. I don't know why VD performs better in SteamVR, but I've added the results below. In VD I used the Ultra setting, which gave me similar resolution to Link at 1.7x, and a bitrate of 150 Mbps (the highest setting VD).

 

Here are my low res results:

 

Rift at 100% (2.15 million pixels per eye):

CV1 100.png

 

Quest 2 at 30% (2.21 million pixels per eye)

AirLink:

Q2 30 al.png

Tethered Link, default bitrate:

Q2 30.png

Tethered Link, increased bitrate (350 Mbps):

Q2 30 hb.png

 

The Quest 2 is pushing slightly more pixels (3% more), but it was the closest I could get with the way SteamVR adjusts resolution. So looking at these results, the Rift does run a little faster, but not anything near 30%. Airlink runs at 94.5% of Rift's framerate, Tethered Link is at 93.0%, and increasing the bitrate dropped it to 92.6%. So overall, a 6 to 7% performance hit when running the Quest 2 at default Rift settings. That doesn't seem too bad.

 

Edit: Virtual Desktop at 30% (2.24 million pixels per eye):

Q2 30 VD.png

Virtual Desktop ran 8.6% faster than Rift, and 14.9% faster than AirLink!

 

 

Here are the high res results:

 

Rift at 344% (7.40 million pixels per eye):

CV1 344.png

 

Quest 2 at 100% (7.40 million pixels per eye)

AirLink:

Q2 100 al.png

Tethered Link, default bitrate:

Q2 100.png

Tethered Link, increased bitrate (350 Mbps):

Q2 100 hb.png

 

The Quest 2 is again pushing slightly more pixels, but there was less difference than the low res tests. Looking at these results, the Rift still runs a little faster, but with even less advantage. Airlink runs at 95.1% of Rift's framerate, Tethered Link is at 98.1%, and increasing the bitrate dropped it to 95.2%. So overall, a 2 to 5% performance hit when running the Quest 2 at maxed settings.

 

Edit: Virtual Desktop at 100% (7.48 million pixels per eye):

Q2 100 VD.png

Virtual Desktop ran 15.5% faster than Rift, and 21.4% faster than AirLink!

 

Obviously results will vary depending on many factors--I'm using a current-generation GPU that may be more efficient at encoding than older cards, and I have more RAM in my system than the PC used for tests posted in the Index thread. Be that as it may, I started using Quest and Quest 2 for PCVR when I was using a GTX 1080, and although I don't have benchmarks for that setup, I never noticed any performance hit that felt like 30% loss.

 

I would be interested in seeing other people's results with different hardware, if anyone cares to do some comparative testing. OpenVR Benchmark is free utility on Steam.

 

On a side note, I haven't used my Rift in quite a while, and OH MY GOD does it look blurry and low-res compared to my Quest 2. Whether comparing them at low or high resolution, the Quest 2 image looks so much cleaner and sharper, while the Rift looks so blurry that I have a hard time reading text in menus, even at 344% super-sampling. It was a great headset in its time, but I could never go back to using it now.

 

TLDR: On my PC, Quest 2 performance overhead ranges from 2% to 7% compared to my Rift CV1, with less performance loss at higher resolution. AirLink versus tethered Link made little difference.

 

EDIT: Virtual Desktop ran 9% to 16% faster than Rift, and 14% to 21% faster than AirLink in SteamVR! I will try to do some testing in a game that uses native Oculus drivers to see how Rift, Link, and Virtual Desktop compare in that environment.

i7 5820K @ 4.25GHz | RTX 3080 Ti FTW3 | Gigabyte GA-X99-UD4 | Corsair DDR4 3000 32GB | Corsair HX 750W
SSDs: Intel 660p M.2 2TB, 3x Samsung Evo 1TB | Startech PCIe 4x USB 3.0 | Startech PCIe 2x USB C 3.1 gen2

30 REPLIES 30

RuneSR2
Level 15

I think we're getting nearer to the real issue here - these results are extremely poor for a RTX 3080 Ti :

 

Quest 2 - RTX 3080 Ti - res 2092 x 2116 = 4.4 mill pixels

Q2 60 al.png

 

- this is one of my old scores from the 3090-thread, got slightly better results the last time, but these old scores still are close enough:

 

Index - RTX 3090 - res 2016 x 2240 = 4.5 mill pixels

RuneSR2_0-1641712078763.png

 

The difference here is a totally massive 41 % !

 

My cpu does have a lot more cores, but this test is normally greatly dependent on the gpu, not the cpu. Also note that refresh rate does not seem to matter much in the OpenVR test - these are other Index results:

 

pr9pgk9d07p51.png

 

With a res of 4.5 mill pixels like the Index, it's normal for at RTX 3090 to get about 80 fps (75 - 85 fps).  

 

Getting 57 fps with Quest 2 using 4.4 mill pixels res with a RTX 3080 Ti is similar performance as Index with an old RTX 2080 Ti. 57 fps would be catastrophic results meaning you've paid many $$$ for a high-end RTX 3080 Ti but you only get RTX 2080 Ti performance. 

 

Also a lot of tests have shown that RTX 3090 and 3080 Ti have same performance, like here:

 

RuneSR2_1-1641713094538.png

https://www.techpowerup.com/review/nvidia-geforce-rtx-3080-ti-founders-edition/28.html

 

Thus the real questions here - why is the Quest 2 performing so poorly? Is it a driver problem? Did Valve limit performance? Are Index drivers really that awesome?

 

For now I'm with @kojack's explanation here: "The thing to remember with an Index vs Quest 2 comparison in an OpenVR benchmark is that the Index is running natively in OpenVR, while the Quest 2 is being forced through an OpenVR to Oculus wrapper written by Valve."

Valve Index & Oculus Rift CV1, Asus Strix OC RTX™ 3090, i9-10900K (5.3Ghz), 32GB 3200MHz, 8TB
"Ask not what VR can do for you, but what you can do for VR"

DaftnDirect
Volunteer Moderator
Volunteer Moderator

Perhaps as Quest use on Steam continues to increase, Valve will take a look at their wrapper and spend more time developing it, with Quest as their primary software sales target. I don't think they're in the business of selling hardware except to insure against others increasing software sales through hardware dependent rival digital stores.

 

Regarding OpenXR vs OpenVR+wrapper... isn't FS2020 OpenXR? if so, that may make a decent indicator of the potential that the wrapper has for optimisation. Perhaps have some Oculus devs write a wrapper for Steam, everyone wins.

 

This is all a reminder that native Oculus support is something to continue to look for in Steam games in the absence of OpenXR inclusion.

Intel 5820K OC@4Ghz, Titan X (Maxwell), 32GB Corsair Vengeance DDR4, ASRock X99 Taichi, Samsung 500Gb 960 Evo M.2, Corsair H100i v2 Cooler, Inateck KTU3FR-4P USB 3 card, Windows 10 Pro v21H2 (19044.2075)

Fully agree - I was wondering why I didn't notice these performance differences before - but my CV1 is totally separated from Steam, I only use my CV1 for games or apps supporting native Oculus drivers.

On my rig, all SteamVR apps and games using Steam drivers always start the Index - and Index is always connected and ready/standby (to minimize any errors of detecting the hmd). 

Thus since I got the Index summer 2019, I have not used any SteamVR games requiring Steam drivers with the CV1.

 

Do support the Oculus Store! I think I now understand what happened when I refunded the Steam version of Trover (which did not support native Oculus drivers) and bought the same game in the Oculus Store - fun to go back to June 2019, where I wrote my Trover review in the Oculus Store - this is a quote:

 

"BTW, initially I bought the SteamVR version, but got it refunded and got the Oculus version. The SteamVR version really let me down today - if I should have rated the game using the SteamVR version, I'd probably rate it 80 % (didn't feel smooth, like semi-glitchy, could easily see the 45 fps x 2 when running). But the Oculus version easily is 95 % to me, it's that good, feels like 90 fps. Or going to the Oculus version felt like upgrading my 1080 to a 2080 Ti, something like that..."

 

I bet the massive performance impact for games requiring Steam drivers has been there for a very long time. Pure coincidence brought it back up when some dude asked me for help as he couldn't understand why his Index was so much faster than his Quest 2 in Steam - and I then asked him to test using the OpenVR Benchmark with the same res.

 

Alyx really felt like a dream come true with the Index and my old GTX 1080, never experienced performance like that before in a Steam game - that's also starting to make some sense if the severe OpenVR Benchmark reductions affect Alyx too. I actually never tried playing Alyx with an Oculus hmd, didn't make much sense using the CV1 instead of Index with that game. 

Valve Index & Oculus Rift CV1, Asus Strix OC RTX™ 3090, i9-10900K (5.3Ghz), 32GB 3200MHz, 8TB
"Ask not what VR can do for you, but what you can do for VR"

kojack
Volunteer Moderator
Volunteer Moderator

It's been so many years and SteamVR STILL can't render CV1 sensors facing the correct direction! They completely screwed up the orientation calculations. Looking at my front sensor, it's aiming down a bit and forwards. The SteamVR view shows it aiming up and backwards.

 

Maybe we should benchmark the Index through Revive, compared to a native Quest 2. 🙂

 

"Regarding OpenXR vs OpenVR+wrapper... isn't FS2020 OpenXR? if so, that may make a decent indicator of the potential that the wrapper has for optimisation."

The only potential issue I see is OpenXR is still early days for most companies. I don't know how much Oculus/Valve/MS have fine tuned performance yet, since Flight Sim is one of the few games using it (only one I know of). But that does sound interesting. I might install Flight Sim again soon.

 

I just spent about two hours playing No Man's Sky on my Reverb G2. I then tested SteamVR with the CV1 to see if the sensors are still rendered wrong as mentioned above. It felt like a vacation! So light and comfortable. The graphics obviously look worse, but not that bad, certainly usable. But it felt so much better to wear it.

 

"Perhaps have some Oculus devs write a wrapper for Steam, everyone wins."

Which is pretty much what OpenComposite is (but not by Oculus). An alternative OpenVR not made by Valve. Although it's still a wrapper around the Oculus runtime, so not ideal.

 

The OpenVR to Oculus wrapper has potential to run fine, but games can do things that the wrapper needs to work around. For example when rendering, Oculus only lets you use a render texture it creates for you, it won't accept a render texture you created (this made working with Ogre3D tricky for me, since Ogre only renders into textures it creates). But OpenVR lets a game choose either way, provide OpenVR with the texture or ask for one. If it does the former, then you have a situation where the wrapper would need to copy the game's texture into the Oculus texture every frame. This is also related to how games like Euro Truck Sim 2 continue to run in DirectX 9 after Oculus moved to DX11 and above: they render in DX9 then copy the texture into the Oculus DX11 provided texture. Not great for performance, but easier than an engine redesign.

 

(Note: above I'm talking about the render texture where the game is rendered in order to pass to the VR system. I'm not talking about textures on models in the scene, that's independent of VR)

Author: Oculus Monitor,  Auto Oculus Touch,  Forum Dark Mode, Phantom Touch Remover,  X-Plane Fixer
Hardware: Threadripper 1950x, MSI Gaming Trio 2080TI, Asrock X399 Taich
Headsets: Wrap 1200VR, DK1, DK2, CV1, Rift-S, GearVR, Go, Quest, Quest 2, Reverb G2

DaftnDirect
Volunteer Moderator
Volunteer Moderator

Interesting stuff.

 

There used to be a lot of conversations like this on the forum way back... maybe I need to spend more time in the developer section now!

 

Yeah, OpenXR is no doubt in its early days and I'm not sure whether the future lies in that or improved OpenVR wrappers. Depends on Valve and developers... and customer hardware choice I guess.

 

Cross-buy and certain Oculus support is still keeping me defaulting to the Oculus store. Steam is my go-to if there's no choice... or if there's a big sale.

Intel 5820K OC@4Ghz, Titan X (Maxwell), 32GB Corsair Vengeance DDR4, ASRock X99 Taichi, Samsung 500Gb 960 Evo M.2, Corsair H100i v2 Cooler, Inateck KTU3FR-4P USB 3 card, Windows 10 Pro v21H2 (19044.2075)

nalex66
Volunteer Moderator
Volunteer Moderator

@RuneSR2 wrote:

I think we're getting nearer to the real issue here - these results are extremely poor for a RTX 3080 Ti :

 

Quest 2 - RTX 3080 Ti - res 2092 x 2116 = 4.4 mill pixels

...57.02 FPS

 

- this is one of my old scores from the 3090-thread, got slightly better results the last time, but these old scores still are close enough:

 

Index - RTX 3090 - res 2016 x 2240 = 4.5 mill pixels

...80.44 FPS

 

The difference here is a totally massive 41 % !


But you'll note, I got exactly the same performance with my Rift, which means it's not the Quest 2 encoding process that's impacting performance, it's Oculus headsets running through Steam (as kojack suggests). That's why I asked if you could test your Rift at a similar resolution with OpenVR Benchmark--because it's misleading to compare your 80 FPS with my 57 FPS and then make conclusions about streaming PCVR. The streaming isn't a factor.

 

Anyway, I think we're all in agreement that we've highlighted a significant performance difference between SteamVR and native Oculus drivers for Oculus headsets. This affirms my decision to buy from the Oculus store rather than Steam when the choice exists.

 

I'm now also satisfied that the performance cost from streaming PCVR to a Quest 2 is nothing to worry about--the difference compared to PC-only headsets seems negligible when comparing apples to apples (at least, on my video card). I'd still love to see some comparisons on lower-tier GPUs to see how they differ in streaming performance, if anyone has a Rift and a Quest 2 that they can comparatively test.

i7 5820K @ 4.25GHz | RTX 3080 Ti FTW3 | Gigabyte GA-X99-UD4 | Corsair DDR4 3000 32GB | Corsair HX 750W
SSDs: Intel 660p M.2 2TB, 3x Samsung Evo 1TB | Startech PCIe 4x USB 3.0 | Startech PCIe 2x USB C 3.1 gen2

kojack
Volunteer Moderator
Volunteer Moderator

@nalex66 wrote:


This affirms my decision to buy from the Oculus store rather than Steam when the choice exists.


Steam does have native Oculus stuff too, but you have to check first to be sure. I know some devs have put oculus sdk support only in the oculus store version and SteamVR in the Steam version (can't remember which games, but I remember seeing it years ago), but I think that's not the standard.

For the real extreme, Dirt on Steam has only Oculus support, no SteamVR. 🙂

 

On the other hand, the Oculus store bans any game it hosts from including SteamVR libraries in the download.

 

 

WMR is so damn annoying. Every time I connect my Reverb G2 I get mouse freezes, display flickering, system near lockup, audio output changing devices, for about 10 seconds or more. A 32 logical core CPU is brought to its knees by the Mixed Reality portal starting up. Oculus software is SO much better.

And the face sensor still doesn't turn off the display when you take the headset off! It sits there getting hotter and burning in the screen until you manually close the Mixed Reality portal app. The only purpose of the face sensor is to change ALL audio to the Reverb (not just VR games) and capture mouse/keyboard input to go into the VR desktop stuff.

 

 

Author: Oculus Monitor,  Auto Oculus Touch,  Forum Dark Mode, Phantom Touch Remover,  X-Plane Fixer
Hardware: Threadripper 1950x, MSI Gaming Trio 2080TI, Asrock X399 Taich
Headsets: Wrap 1200VR, DK1, DK2, CV1, Rift-S, GearVR, Go, Quest, Quest 2, Reverb G2

DaftnDirect
Volunteer Moderator
Volunteer Moderator

I thought we had a list of native Oculus games on Steam at one point but I can't find it so maybe I'm imagining it.

@nalex66 I'll try and do a Rift-S/Quest 2 comparison when I get a chance, am away form home again for a week though so am enjoying standing-alone.

 

Edit: incidentally I'm re-ripping my small collection of 3D blurays to watch on the Quest 2, they look so good, way better than my TV, active 3D on TVs was pretty dire imo. There's not too many true 3D movies though, some seem to be converted 2D.

Intel 5820K OC@4Ghz, Titan X (Maxwell), 32GB Corsair Vengeance DDR4, ASRock X99 Taichi, Samsung 500Gb 960 Evo M.2, Corsair H100i v2 Cooler, Inateck KTU3FR-4P USB 3 card, Windows 10 Pro v21H2 (19044.2075)

If there's a list, it certainly needs to be updated. To find out if a game has native Oculus driver support it may be slightly cumbersome. For some games it's hidden - like my new game The Shore VR. 

For some games, when visiting my Steam Library and pressing the start button for a game, I will be given a choice to use SteamVR or Oculus Mode. So that's easy and straightforward. Might only need to be done once, after that (like Hinge) the game will be picked up by the Oculus desktop app and registered in your Oculus library. 

A game like The Shore VR does not give you any options - but when you click the play button the game automatically detects your hmd - and lo and behold it started in my CV1, because I was wearing the the CV1. If I wear the Index, then The Shore VR will automatically start the Index. But many times I've worn the CV1 hoping that a game would start only to hear my base stations starting - the sound of CV1 doom, lol. 

 

In short, for many games I need to buy them and then hold my breath while using the CV1 and starting the game - often it's fine just clicking the game's desktop icon. But for some games you may need to start them from the Steam Library and chose the Oculus Mode. 

Most games will integrate just fine in the Oculus desktop app - but recently I'm getting some really ugly icons - earlier I got nice game icons in the Oculus desktop app, maybe Steam changed something... 

 

If everything fails, you can try to ask the game devs for Oculus support - I was down on my knees here for Ancient Dungeon (my son only gets to use the CV1, he can't play with no Oculus support):

 

https://steamcommunity.com/app/1125240/discussions/0/3156454576343994002/

 

The devs provided Oculus driver support a few weeks ago - turns out my 9 year old son got scared of that game, but he likes playing Tempestfall - which I consider much scarier. Tempestfall has native Oculus driver support too. 

Valve Index & Oculus Rift CV1, Asus Strix OC RTX™ 3090, i9-10900K (5.3Ghz), 32GB 3200MHz, 8TB
"Ask not what VR can do for you, but what you can do for VR"

TomCgcmfc
Level 15

@nalex66   Good starting post mate and, in general, I agree with your findings.  All these benchmarks give indicative results, not definitive though imho. 

i9 9900k, rtx3090, 32 Gb ram, 1tb ssd, 4tb hdd. xi hero wifi mb, 750w psu, QPro w/Air Link, Vive Pro