cancel
Showing results for 
Search instead for 
Did you mean: 

UPDATE: Qualcomm Statement [reversed] Suggesting Quest 2 Sales Numbers

kevinw729
Honored Visionary

kevinw729_0-1637145084021.png

 




https://usa-sciencenews.com/2021/11/16/metas-oculus-quest-2-has-shipped-10-million-units-according-t...

https://vrawards.aixr.org/ "The Out-of-Home Immersive Entertainment Frontier: Expanding Interactive Boundaries in Leisure Facilities" https://www.amazon.co.uk/Out-Home-Immersive-Entertainment-Frontier/dp/1472426959
26 REPLIES 26

I too find it hard to believe that Qualcomm got these numbers wrong. I'm confindent that Qualcomm has indeed shipped 10 mill XR2 gpus to Facebook, even if Facebook for reasons unknown may not want that number out in the open. And of course sold units may be less - some estimate about 7 mill. In order not to experience chip shortages, it may be a smart move for Facebook to stock up on XR2s... Even if these are already slow with the old Adreno 650 gpu, Snapdragon 888 with Adreno 660 is about 30% faster. So you can already get new phones wihich are much faster than the outdated Quest 2 with Adreno 650, lol. 

 

Btw, I wonder if Cambria will use the Snapdragon 888 - and the Adreno 660 - but it will still be nothing more than dust beneath the feet of even an old GTX 970.

 

Snapdragon_888_adreno_660.png

 

For tflops (FP32) we get:

 

Quest 2 (XR2, Adreno 650) = 1.2

Snapdragon 888 (Adreno 660) = 1.7

GTX 970 = 3.9

GTX 1080 = 8.9

Playstation 5 (AMD Oberon gpu) = 10.3

RTX 3080 = 29.8

 

- or we could set the 3080 to 100% - then we get:

 

RTX 3080 = 100%

Playstation 5 (AMD Oberon gpu) = 35% (RTX 3080 is 3 (2.9) times faster)

GTX 1080 = 30% (RTX 3080 is 3 (3.3) times faster)

GTX 970 = 13% (RTX 3080 is 8 (7.7) times faster)

Snapdragon 888 (Adreno 660) = 6% (RTX 3080 is 17 (16.7) times faster)

Quest 2 (XR2, Adreno 650) = 4% (RTX 3080 is 25 times faster)

 

So all that Qualcomm makes is dust compared to real PCVR and even the upcoming PSVR2. Even more if Nvidia Series 40 arrives next year. 

Oculus Rift CV1, Valve Index & PSVR2, Asus Strix OC RTX™ 3090, i9-10900K (5.3Ghz), 32GB 3200MHz, 16TB SSD
"Ask not what VR can do for you, but what you can do for VR"


@RuneSR2 wrote:

For tflops (FP32) we get:

 

Quest 2 (XR2, Adreno 650) = 1.2

Snapdragon 888 (Adreno 660) = 1.7

GTX 970 = 3.9

GTX 1080 = 8.9

Playstation 5 (AMD Oberon gpu) = 10.3

RTX 3080 = 29.8

 

- or we could set the 3080 to 100% - then we get:

 

RTX 3080 = 100%

Playstation 5 (AMD Oberon gpu) = 35% (RTX 3080 is 3 (2.9) times faster)

GTX 1080 = 30% (RTX 3080 is 3 (3.3) times faster)

GTX 970 = 13% (RTX 3080 is 8 (7.7) times faster)

Snapdragon 888 (Adreno 660) = 6% (RTX 3080 is 17 (16.7) times faster)

Quest 2 (XR2, Adreno 650) = 4% (RTX 3080 is 25 times faster)

 

So all that Qualcomm makes is dust compared to real PCVR and even the upcoming PSVR2. Even more if Nvidia Series 40 arrives next year. 


 

That is not an accurate comparison. When we use Quest for a PCVR headset, it gets the same benefits from an RTX 3080 as any other HMD on the market. Yet with Quest we also get the tetherless experience, so it can arguably be said that Quest offers a superior (real) PCVR experience compared to other units that are forced to remain tethered.

 

I am not sure why someone would compare Qualcomm to a GPU when we know that Quest can connect to a PC for PCVR and take full advantage of the GPU.


Granted, there is the whole "native vs streaming" argument around the GPUs roles in PCVR, but the differences are becoming more and more trivial as this technology advances.

Btw, also good to hear from you again - hope all is well! 

 

Forgot to mention that comparing tflops (FP32) may make the Quest 2 (XR2) look way too good - for example GTX 1060 has 4.4 tflops, and with 1.2 tflops one could think that Quest 2 is close to 25% of the power of a GTX 1060 - but tflops is just a synthetic number and getting closer to the real world, GTX 1060 is about 6 times faster than the Quest 2 (GFXBench):

 

OculusHeadsetGPUsCompared_PC2.png

https://uploadvr.com/oculus-quest-2-benchmarks/

 

Of course Quest 2 with its ultra-low-end Adreno 650 phone gpu is nothing more than dust compared to real PCVR, and games for locomotives should not be mixed with games for small cars in the Award show, lol. And RTX 3090 is about 5 times faster than the GTX 1060 - so that's about 35,000 points? 😎

Oculus Rift CV1, Valve Index & PSVR2, Asus Strix OC RTX™ 3090, i9-10900K (5.3Ghz), 32GB 3200MHz, 16TB SSD
"Ask not what VR can do for you, but what you can do for VR"

nalex66
MVP
MVP

And yet, this headset (that is apparently dust-powered) is providing me an extremely enjoyable experience playing through the native version of Medal of Honor. You know, MOH, the game that some PC players will declare needs a top-end video card to be enjoyed. Somehow, despite a woeful lack of synthetic benchmark points, the experience of fighting across war-torn Europe feels remarkably similar to the PC version. Granted, the visuals don’t have the same fidelity, particularly in large outdoor levels. Even so, the gunplay feels good, the campaign manages to tell its story, and the game is fun. It’s almost like all this talk of teraflops, and what’s faster than what, is a distraction from what actually matters, which is the experience provided to the user.

 

Oculus hasn’t sold millions of Quest 2 headsets because it provides a crappy VR experience. These things are flying off the shelves because they’re awesome. As usual, the people who are most critical of the Quest’s capabilities are the ones who haven’t actually tried it. I’ve seen many testimonials from people who’ve given it a chance and were amazed by what the Quest 2 can do, both as a stand-alone headset, and as a PCVR upgrade. If everyone was so dismissive of something they’ve never experienced, VR itself would never have taken off.

DK2, CV1, Go, Quest, Quest 2, Quest 3.


Try my game: Cyclops Island Demo

- and I may have posted in the wrong thread, sorry. My point was, directed at the VR Award Show thread, where @Anonymous mentioned the issue of mixing mobile games and PCVR games that I also believe it would be fair to separate mobile VR and PCVR due to the differences in processing capabilities - also devs are working under very different conditions making games for PC and mobile processors like the XR2. 

Oculus Rift CV1, Valve Index & PSVR2, Asus Strix OC RTX™ 3090, i9-10900K (5.3Ghz), 32GB 3200MHz, 16TB SSD
"Ask not what VR can do for you, but what you can do for VR"

> My point was, directed at the VR Award Show thread, where @snowdog mentioned the issue of mixing mobile games and PCVR games that I also believe it would be fair to separate mobile VR and PCVR due to the differences in processing capabilities

 

I think it is not so simple. I think better to separate "processing power comparison for VR" from "VR feature usage" comparison. And yet another "new VR features/usage" (unlike any other "game" before, like "detect knock on headset", "Touch as marker", "Touch on legs", "Touch on kids toy"....).

 

So for processing power comparison you could split in more then 2 categories. It could be arm processors (both for mobile and stationary devices if there are any (future Steam Deck? may be)), low end PC gaming cards (old?), top end gaming cards etc.

 

And for features there could be other categories like "jumping in VR with real life jump" or "laying on the floor". Or "could you grip to other avatar". Or even "is there force feedback suite integrated in game"/"touch gloves" or some "smell generating device". So you could count how much independent VR features there are.

 


@RuneSR2 wrote:

Of course Quest 2 with its ultra-low-end Adreno 650 phone gpu is nothing more than dust compared to real PCVR


 

Nope. This statement mixes technologies incorrectly.

 

The term "PCVR" refers to the software platform supporting a VR HMD. Quest can very much run PCVR titles just as well, if not better, than something like the Index, Reverb, Pimax, etc.

 

Quest 2 can connect to a PC both wired and wirelessly and take advantage of the PC's GPU.

 

Quest 2 can be powered by an RTX 3090.

 

The difference between something like Quest 2 and an Index... is that Quest 2 can take advantage of a 3090 while being 100% tetherless. On top of the fact that something like the Index has worse SDE and Glare than Quest 2; thus giving Quest 2 an advantage visually since the image in a Quest 2 being powered by an RTX 3090 will be much clearer than the image in an Index using the same graphics card. 

 

Therefore, both visually and convenience (tetherless inside-out tracking), it would be more accurate to say that Index is nothing more than dust compared to Quest 2 (if we are forced to use your phrasing here).

 

Which is probably why Valve hasn't released any info on an Index CV2 and instead has taken out a patent on their own Stand-Alone HMD. 😎