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

29 REPLIES 29

"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."

 

Fully agree - and your work showing limited impact of steaming is indeed very important. To me it makes Cambria much more attractive if that hmd uses streaming like the Quest 2, especially if the impact of streaming is about 5% (average), and maybe down to only 2% for the high-res results. 

 

I do find it very problematic with the Steam performance - and it seems HP (WMR) hmds are impacted too judging from Kojack's test results, so might not be a performance issue limited to Oculus hmds. Could indicate a performance reduction of about 30% for all hmds not using native Steam drivers - and only Index and some HTC Vive hmds use native Steam drivers (Tom got nice results when using his Vive Pro in OpenVR Benchmark too - 76 fps in 2016x2240 if I remember correctly). Valve must know about this performance difference, but it's very easy to hide for consumers who use many different hmds and settings. To show the difference you have to set the res similarly for two different hmds and test the same game/app - and since most games run at 45 or 90 fps, it's hard to measure directly in a game and to get precise results. 

So evidence is hard to dig up - but this dude did some nice work and ran Quest 2 vs. Vive using similar res in FallOut 4 (=a Steam game with no support for native Oculus drivers) and confirmed that the Vive performance was extremely better, Quest 2 got about 50-60 fps, while Vive got 70-80 fps, which fit the OpenVR numbers of the Vive being about 30% faster at same res (both hmds used 6.5 mill pixels res in the FallOut 4 test) - go to 10:55 in this vid: https://youtu.be/4Ibiaj5sio0 (The title on that vid isn't good - when testing native Oculus games results should be much better for the Quest 2. Actually the dude did get nice results in Tales of Glory, and that Steam game supports native Oculus drivers, but obviously the dude did not know that, lol)

 

If Valve more or less deliberately is reducing performance (or simply not doing anything to improve/optimize performance) for hmds not using native Steam drivers, this is a great problem. Might of course affect the enjoyment of VR, especially if more than 60% (maybe even 70% including WMR and more) are getting subpar performance when running VR games requiring Steam drivers. Might post this on Reddit later to see if there're more perspectives on this. 

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"

nalex66
Volunteer Moderator
Volunteer Moderator

I wonder how much OpenComposite might improve things in Fallout4VR? I haven't messed around with it, but some people who've had trouble with SteamVR swear by it. I tried it to see if OpenVR Benchmark would work with OpenComposite, but that was a no-go.

 

I did test Fallout4VR when I got my new video card, and it seemed to run well (and looked absolutely stunning) with my maxed settings (5408x2732), so I didn't worry about performance.

 

Last time I tried No Man's Sky, it still had performance issues, but I don't remember whether I tried it with the RTX 3080 Ti. That game might be a good use case for OpenComposite, if it improves performance over SteamVR. That said, I rely on SteamInput to rebind some awkward controls in NMS, and presumably that won't work without SteamVR running.

 

Hopefullly as we move towards an OpenXR standard, these performance discrepancies can be improved or eliminated.

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

If I had an Index or Vive, I'd investigate what's actually going on. But the Index still isn't available here (it's $1899au, but out of stock at EBGames (the only authorized reseller). I've never seen it in stock) and I can't stand using Vives.

I guess when I'm back at work I can use the Vives there, but my workload has gone up, last thing I want to do on campus is hang around after work using a Vive.

 

 

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

nalex66
Volunteer Moderator
Volunteer Moderator

I had a request from Reddit to benchmark Virtual Desktop too. I really didn't expect its performance to differ from AirLink, but I went ahead and did the tests. To my surprise, those results exceeded both Link and Rift. I don't know why VD performs better in SteamVR, but I've added the results alongside my other benchmarks above. 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).

 

My Virtual Desktop results:

 

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!

 

Virtual Desktop at 60% (4.48 million pixels):

Q2 60 VD.png

 

Virtual Desktop ran 16.0% faster than Rift, and 14.4% faster than AirLink!

 

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!

 

I was so surprised by these results that I repeated my AirLink tests, just to make sure nothing else had changed, but I got similar results to my previous benchmarks. I haven't used VD much since AirLink came out, but I suppose I should reconsider that in light of these results.

 

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

Fantastic work, @nalex66 !

 

This thread should be stickied for a time, as it truly has invaluable information that will help consumers not only with purchasing decisions, but also with configuration management and troubleshooting performance issues!

 

Thanks to the info provided by you and @kojack, VR Enthusiasts investing in Oculus VR (oops, Meta VR) can start baselining their benchmarks much more accurately and efficiently.

 

Cheers all!

kojack
Volunteer Moderator
Volunteer Moderator

Ok, those VD numbers are making more sense of the situation.

 

To my understanding Virtual Desktop, when used with SteamVR, bypasses the PC Oculus SDK so doesn't use the Valve written wrapper. It acts as the wrapper, directly streaming the OpenVR frames to the headset where the VD Quest app receives them and talks to the Quest Oculus SDK.

 

Although I've only used VD a couple of times. But I'm pretty sure that's what it is doing based on what I've read.

 

Of course I'm not saying Valve are doing anything on purpose, they aren't Nvidia.

Nvidia have been busted disabling features if AMD hardware is present (hardware PhysX not allowed if any AMD card present, even if a Nvidia card is there too) and detecting benchmark software from executable name so they can change rendering to get better results. I've tested the latter myself about 15 years ago, renaming a game exe to match a well known benchmark gave a higher frame rate because the drivers started cheating on settings.

 

We definitely need an Oculus native benchmark, to test VD on Oculus vs Airlink.

 

I tried OpenComposite with OpenVR benchmark, yeah, it doesn't work. But OpenComposite is open source (and not updated since 2020), maybe I can see where its failing...

 

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

kojack
Volunteer Moderator
Volunteer Moderator

I got OpenComposite / OpenVR Benchmark to the point where most of the bits were fixed (had to add some bits and bypass some safety checks), but now it crashes in Unreal. Apparently OpenVR Benchmark is an Overlay App instead of a Scene App, and OpenComposite doesn't support Overlay Apps (ones that can draw on top of other VR apps). Why is OpenVR Benchmark made like that? Does it actually use overlay functionality?

 

So I'm at the end of my quick look into OpenComposite fixing for now.

 

(Oh yeah, OpenComposite was updated more recently than 2020, it's moved to OpenXR support 5 months ago. I'm compiling that branch, it's the most recent)

 

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

Interesting VD results indeed @nalex66 - and I think Valve is just lazy as ever. Normal users should not have to worry about cables, airlink, VD etc., (and I'm pretty sure most do not even know there may be great performance differences) - they should just be able to use whatever hmd they have - which is compatible with Steam - and get optimal performance. Sadly this does not seem very realistic at the moment. 

My bet would be that most common Quest 2 users choose Airlink or cable, few will probably use VD (which isn't free, but costs $20 in the Quest Store). 

 

65 fps at 4.5 mill pixel res are definitely much better than 57 fps. It's not just 8 fps, 65 fps are 14% faster than 57 fps - in the current gpu world, 14% extra performance can be very expensive.

 

65 fps at 4.5 mill pixel res are close to low 3080 (non-Ti) results for Index and Vive Pro users (the 3080 results I've seen range from 66 to 72 fps for Index and Vive Pro). You should be able to get at least 75 - 85 fps in a perfect world with the 3080 Ti, but 65 fps are indeed a big step forward!

 

PCVR isn't doing that well at the moment, and PCVR has always been very hardware demanding - the last thing we need is tons of hmds getting suboptimal performance using SteamVR due to bad drivers, it should be in Valve's best interest to secure that all users get the best possible experience - but the VR software sales on Steam may be a drop in the ocean compared to pancake gaming, and thus too little for Valve to care... 

 

Btw - the Reddit discussion is here and does contain many interesting comments:

 

https://www.reddit.com/r/oculus/comments/s0t17e/performance_overhead_due_to_linkairlink_it_may/

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"

nalex66
Volunteer Moderator
Volunteer Moderator

@kojack wrote:

 

We definitely need an Oculus native benchmark, to test VD on Oculus vs Airlink.


I’ve been thinking about how best to benchmark under the Oculus API. The Virtual Desktop results are fascinating, but they’re quite specific to SteamVR (or even just this particular benchmark tool). I’m curious to see how Rift, Link, and Virtual Desktop compare in a native Oculus title.

 

Perhaps I should just load up Stormland or Asgard’s Wrath, max out the settings, and just stand in one spot looking at some specific target. I can use the Oculus Debug Tool to bump up the Rift to match the Link and VD resolution, and the performance HUD to check my frame rates. It might be hard to get meaningful results though, if I’m still hitting 90 fps. I guess I’d also have to disable ASW.

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

That's how I use to do it - increase (or decrease) res to find the threshold for 90fps, find a demanding place in a game and don't move 😉 This may be easy with the Oculus Tray Tool controlling global ss - many games, at least with CV1, will instantly use the new ss values. But with OTT I need to change HUDs many times to see performance and verify pixel densities - very cumbersome. Using Steam I often have to restart a game for new SteamVR res settings to take effect - also very cumbersome (and don't trust fpsVR in Steam - that program just reads the SteamVR res value even if the value hasn't been applied to the game before a game restart). 

I guess that's why we rarely see benchmarks in VR, it's very time consuming. 

Some games have build-in fps counters, like Xing The Land Beyond and Tempestfall, but it's still a hassle. Testing real games is the best way of course, as synthetic benchmarks aren't real games - but at least the OpenVR Benchmark is using the game engine from the game CyubeVR.

 

BabelTech has written a nice article about benchmarking VR games - and uses special equipment and software:

 

https://babeltechreviews.com/oculus-rift-vr-benching-amd-vs-nvidia-part-2/

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"