Alongside the announcement of the new Snapdragon XR2+ Gen 2 chip, Qualcomm also revealed its latest reference headset, a mixed reality device that’s likely the foundation of Samsung’s upcoming headset.

Samsung announced last year that it’s working with both Google and Qualcomm on an XR device. We’ve been around the block once or twice and can tell you how these things usually go down.

Qualcomm is far and away the leading provider of processors powering XR devices today. Part of the reason why is because when the company makes a new XR chip, it makes more than just the chip.

Qualcomm also creates ‘reference headsets’, which are bare-bones but fully-functional headsets that contain their latest chip, and act as a blueprint for customers (ie: Meta, HTC, or Samsung) to use as a starting point to develop their next device.

Lo and behold, alongside the Snapdragon XR2+ Gen 2 announcement this week, Qualcomm not only confirmed that Samsung would be using the new chip but also revealed a new MR headset reference design which very likely gives us a ballpark idea of what Samsung will reveal.

Image courtesy Qualcomm

Like most of Qualcomm’s reference headsets, the Snapdragon XR2+ Gen 2 reference headset is made in partnership with Goertek. And here’s what we know about it:

  • Snapdragon XR2+ Gen 2 processor (obviously)
  • Eye-tracking from Tobii
  • Support for 18.5MP (4,300 × 4,300) per-eye resolution at 90Hz
  • 12 concurrent cameras
    • [Our speculation on their uses]
      • 2x eye-tracking
      • 2x RGB passthrough
      • 4x world-tracking
      • 4x depth-sensing
  • Pancake lenses
  • Hardware IPD adjustment
  • Microphone array
  • 3.5mm headphone port
  • Wi-Fi 6/6E/7

So that’s the reference headset. What can it tell us about what Samsung is likely building?

Well for one, the overall form-factor. While Samsung’s headset will have different aesthetics, the form-factor probably won’t deviate all that much. That’s because the XR2+ Gen 2 is built for a certain amount of power consumption, heat dissipation, and camera count. If Samsung was building something significantly smaller, they would likely be building with the Snapdragon AR2 platform instead, which is designed to be more power-efficient and compact.

The camera stack is likely to be similar too, considering that XR2+ Gen 2 ups the concurrent camera count to 12 (compared to 10 with the prior XR2 Gen 2).

But then there’s the parts that are usually heavily customized between the reference headset and the finished product. The first of which is the display.

While the reference headset can technically manage up to 18.5MP (4,300 × 4,300) per-eye resolution at 90Hz, our understanding is that Qualcomm’s reference headsets usually aren’t equipped with the maximum supported display (because the maximum theoretical pixel throughput usually exceeds displays that are widely available at the time).

SEE ALSO
'Fisherman's Tale' Studio Unveils Mysterious VR Adventure 'One True Path' for Quest 3 & PC VR, Trailer Here

The status quo for displays of this size currently on the market is in the 2K range. Samsung’s headset will likely land between 2K and 3K per-eye; we doubt they’ll jump all the way to 4K.

It remains to be seen if Samsung will stick with the industry-standard LCD display technology or perhaps want to leverage its own OLED display expertise. Samsung acquired a microdisplay manufacturer last year, though it isn’t clear if that was to support this specific upcoming headset or simply a strategic move anticipating future demand for such displays.

The lenses are another element that tends to differ significantly from the reference design. Recent headsets have been moving away from single-element Fresnel lenses and toward more compact pancake lenses. Samsung’s headset will likely use pancake lenses but we’ll have to wait and see if they have something as good as those in Quest 3, or if they’ll end up with something closer to the less performant lenses in Vive XR Elite.

The big question up in the air right now is whether Samsung’s headset will be positioned to compete with the likes of Meta’s Quest or Apple’s Vision Pro; our bet is on the former.

Image courtesy Qualcomm

Interestingly, Qualcomm’s marketing photo shows the user with her hand in her lap doing a pinching gesture. That looks an awful lot like the look-and-tap system that Apple has shown on Vision Pro. Will this become a new standard input method?

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.


Ben Lang
Ben is the world's most senior professional analyst solely dedicated to the XR industry, having founded Road to VR in 2011—a year before the Oculus Kickstarter sparked a resurgence that led to the modern XR landscape. He has authored more than 3,000 articles chronicling the evolution of the XR industry over more than a decade. With that unique perspective, Ben has been consistently recognized as one of the most influential voices in XR, giving keynotes and joining panel and podcast discussions at key industry events. He is a self-described "journalist and analyst, not evangelist."
  • Andrew Jakobs

    Let’s hope Samsung announces a new headset at the galaxy unwrap event half january. Hopefully finally with oled panels and at least the same good lenses of the Q3, resolution also a small bump, with eyetracking for automated IPD, and for a price between the Q3 and Q pro. But WITH decent controllers. If they are about the same as Q3 but faster SOC it should be around the Q3 or lower.

  • Stealth Ico

    Funny to see everyone jumping on the XR train now that Apple committed.

    Hopefully they all have good software, otherwise they will just be handing Apple the win at the end.

    • Garhert

      I’m confused… is there a new XR train now because Apple is releasing a product?

      • Stealth Ico

        Yeah. In the eyes of normies, Apple jumping into the fray kind of legitimizes the space.

        • Garhert

          Normies, yes, but I don’t see any (new) OEM jumping on this train just because of Apple…?

          • ApocalypseShadow

            You mean the fact that Facebook dropped their conference right next to Apple’s announcement? The fact that they are positioning their headset as a low cost Vision Pro? The fact that Bytedance is pivoting from VR to AR with use of hand tracking with Pico? The fact that Samsung, in collaboration with Google, went back to the drawing board to make sure their headset competes with Apple?

            There’s more headsets coming. And as much as some think that Apple isn’t a threat because the launch headset is $3,500. The simple fact is that Apple is able to sell $1,000 plus iPhones without a problem because of its symbol of status. While Facebook couldn’t sell a $1,500 Quest Pro and sales of Quest 3 at $500 are way slower than Quest 2 warehouse clearance pricing to make room for their next lite model with the new chip built in that’s in the Quest 3.

            Apple having such a huge platform with cellular backing is something Facebook wants. And they know Google with Samsung, are the other cellular companies that are threats to their monopolizing the mixed reality space.

          • Garhert

            This is not “jumping on the XR train”. The train has been on track for years, Apple may just accelerate it. They do not possess the Holy Grail as some believe.

          • Andrew Jakobs

            Pico already showed handtracking with their Pico 4 Pro, so don’t try to make it if Apple is the one who came up with it, nothing about the AVP is something we haven’t seen before.

          • Christian Schildwaechter

            Nobody actually knows what Apple came up with for AVP, as the project has been run for a decade in total secrecy. And “XYZ did it first” is a silly way to judge technological progress, as that is usually someone in a lab, a company adding immature tech as a sales argument or something else without lasting impact. It often takes a lot of time to turn something from proof of concept into a useful tool, and neither Pico nor Meta so far managed to make hand tracking a viable alternative as the sole/main way to use their headsets.

            Interestingly only those who despise Apple keep claiming that Apple fans/users believe Apple invented the first mouse, mobile music player, multitouch screen, smartphone etc. What Apple actually did was wait for the basic technology to mature enough, and only then implement it in an intuitive and user friendly fashion that often became the standard way of doing things. Nobody really cares about tech that isn’t usable, and where Apple regularly scores is being the first to implement tech in a way that isn’t barely usable rubbish.

      • Dave

        Garhert you are not keeping up with the news son.

        Qualcomm announced 5 clients already activily building headsets using this very chip. I’m pretty confident all of these will have a strong XR capability.

  • Bob

    “The big question up in the air right now is whether Samsung’s headset will be positioned to compete with the likes of Meta’s Quest or Apple’s Vision Pro; our bet is on the former.”

    Not too sure about that. If I recall, Samsung delayed their XR product from last year due to the annoucement of the Vision Pro. Clearly, if they intend to launch a product this year, it has been designed and engineered to be on par with the Vision Pro in terms of specifications which would warrant a significant price tag a la $3000+.

    Expect a super expensive headset designed and marketed to pull the tech enthuisiasts with loaded pockets by the end of the year.

    • Dave

      I don’t see how this can compete with the Apple Vision Pro. The XR2+ Gen2 architecturally is from the XR2 family, it doesn’t have a dedicated passthough processing node, so how can it accomplish the things which the Apple Vision Pro does. This headset will end up being closer to the Quest 3 than Vision Pro but it’s nice to have Ultraleap handtracking support, so that will be decent as well. Should offer a great experience though but I don’t see it being ‘competitive’ to the Vision Pro.

      • Cl

        What benefits does this have? From what I’ve seen the only difference is that avp has higher resolution passthrough, which could have been done on quest3 if they used better cameras.

        • Dave

          Yes so Apple Vision Pro use a secondary processing chip R1 to stream images to the display in 12 ms. The camera and depth sense computations mitigate the warping you see in the image when viewing objects at various distances. Something the Quest 3 really struggles with. I’m sure an XR2+ Gen2 will be better at this, but it’s unlikely to provide the sort of solutions for image passthrough processing nor the quality of sensors Apple are using in the Vision Pro. We’ll see, I hope I’m wrong but I can’t see the Samsung/Google headset overcoming these issues unless they are using significantly better cameras and have a super efficient software pipeline for handling the passthrough processing.

          • Christian Schildwaechter

            XR2+ Gen 2 actually provides very low latency, less than 12ms video pass-through according to Qualcomm, or one frame of latency at 90Hz. A while ago someone tested the passthrough latency on XR2 Gen 1 headsets, and came to 40ms on Lynx R1 that just passes through the image, and 50ms on Quest Pro that first processes it to reduce distortions. The 12ms on XR2+ Gen 2 probably refer to unprocessed passthrough, but that’s still a very significant reduction by 70% compared to XR2 Gen 1 on Lynx R1.

            We’ll have to see how fast/well the full pipeline works and how it compares to AVP, but the massive compute power of AVP is not needed for the “spatial” display or hand and eye tracking, which by now are surprisingly “cheap”. Motion prediction for ETFR might still be very computationally expensive, but the latest numbers we got were for Tobii tracking with the HTC Vive Pro Eye, so this may have improved too. For simple apps like virtual cinema, an XR2+ Gen 2 HMD could provide comparable performance to an AVP. Until a lot more XR software has been developed for AVP, the main use of all that CPU power will be running desktop class iPad apps like Final Cut or Logic, which on current iPad Pros fully utilize its M2.

    • g-man

      Note the woman in the picture here too, doing an AVP-style low-down pinch gesture. I doubt that’s a coincidence.

  • MackRogers

    Qualcomm is the worst company in the world.

    All their benchmarks are never under load.

    The reason why Meta is a joke is because they thought Qualcomm had their best interests in mind. Qualcomm has been trying to stifle innovation and slow roll their chips for years.

    Should have went custom silicon for the 36 billion you lit on fire ZUCK. Now look at you. Clown.

    • ViRGiN

      You could have gotten job as personal Mark advisor; instead you are giving out all the secret info out there for free in a comment section. Good work champ!

    • Christian Schildwaechter

      Qualcomm not being nice to their partners/customers and relying on a lot of shady methods/blackmail is well known, so it’s very unlikely that Meta ever thought Qualcomm had their best interest in mind. And Meta tried to develop their own custom silicon for years, most likely spending billions, hoping to evade dependency on Qualcomm’s (often rather lacking) willingness to improve technology and their (never lacking) willingness to exploit and extort and resulting high costs.

      But creating competitive ARM silicon is f*%$ing hard. Samsung for years tried to replace high end Qualcomm SoC with their own Exynos featuring proprietary “M” cores, but always struggled and in the end switched back to designs licensed from ARM. Qualcomm has been struggling to catch up to Apple with their modified ARM cores, and finally bought Nuvia in 2021 for their advanced core designs, which will become the new and improved Oryon core, probably starting with SD8 Gen 4. And Apple themselves bought chip designer P.A. Semi in 2008, who had released very power efficient designs, solely for their team with decades of experience, to help with creating their first own silicon with the A4 in 2010. And it still took more than a decade to come up with something they could compete with/beat Intel’s mobile chips.

      “Should have went custom silicon” is a bit like suggesting someone should found their own rocket company as an alternative to SpaceX, because they think Elon Musk is a dangerous looney. Nice idea in theory, but really hard and expensive in reality, taking a long time with no guarantee of success. A couple of billionaires tried, and so far the main result is them having fewer billions.

      • MackRogers

        Nobody indicted it was easy or likely but it is well known their FAST team never had a chance and was set up to fail. It was undermanned and lacking recruited talent. You don’t just dip your toe in the water for custom silicon.

        This was a leadership issue and lack of priority on the part of Zuck and Boz. They didn’t put the time or work in and are now perennially behind Apple by orders of magnitude.

        • Christian Schildwaechter

          I don’t know what their considerations were, but with MRL burning USD 10bn/year and having to convince investors that this will pay off in the far future, pushing an extremely costly and very complex long term hardware project may have not been worth the risk for limited benefits. Beating Apple, who distribute their SoC development costs over hundreds of millions of iPhones, Macs and other devices sold every year, or Qualcomm, who pretty much power all high end Android phones, would have been both unlikely and ridiculously expensive for a few million HMDs.

          I always considered that an insane idea with no valid business case, esp. since Meta never wanted to be a hardware company. They only produce the Quest to bootstrap XR/the metaverse as a platform, where they intend to make they money back. So creating their own silicon would be rather pointless in the long term, esp. since VR hardware is already pretty much commoditized similar to PCs, to reduce development cost in the very tiny VR market.

          Money is much better spend on XR specific hardware and software, with as many components as possible taken from existing platforms, like CPUs, GPUs, cameras, batteries, RAM, storage etc., which is way cheaper. It makes sense to develop low power neural chips that can be integrated into ARM SoCs for rendering photorealistic avatars, much more than developing a completely new ARM SoC from scratch that will do pretty much the same as existing ones.

          • MackRogers

            People smarter than you and I at Meta looked at the landscape and apparently thought it was feasible which is why they created a 600 engineer FAST team and invested Billions. So, it’s not just me who thought it was a good idea. Apparently they did too. Zuck just underfunded the program and didn’t recruit talent to make it a reality and wanted the stock pump from pulling the plug in “the year of efficiency”(to his credit he got it, because wallstreet are morons). Just grab an off the shelf Snapdragon right?… wrong. No look at them, at Qualcomm’s mercy.

            Secondly, you make the strawman argument of Apple’s history of bionic chips and there is no way they could catch up to 20 years of engineering.

            They don’t have to, look at Google Tensor it’s not a dramatic redesign it’s a chip that is custom TO THEIR NEEDS. Google isn’t re-inventing the wheel, they are prioritizing efficiency and TDP.

          • Christian Schildwaechter

            MRL’s strategy of starting lots of projects and see what sticks lead to a series of inconsistent, universally hated VR social spaces. Them willing to spend money on a project doesn’t mean it isn’t stupid, as in “makes no business sense, even if technically feasible with enough funding.”

            As mentioned above, creating specialized/optimized components for SoCs makes a lot of sense, and is what Google did with Tensor and Meta plans with Qualcomm for NPU. ARM SoCs can be configured for different use cases, and Google Tensor still uses CPU cores licensed from ARM. ARM is now trying to push new licenses that also require the use of ARM designed GPUs and NPUs, which could force Google to move the tensor cores to a separate chip in the future.

            While Google added a garage, Apple rebuild the house. Creating your own silicon/CPU requires an expensive ARM architecture license that Apple, Qualcomm and Samsung have, but Google doesn’t. The M1 was made a feasible SoC for running existing software by extending the CPU architecture with x86 instruction flag handling that other ARM CPUs have to slowly emulate, causing Windows on ARM to sucks. This and selling billions of vertically integrated devices made custom silicon a sound idea despite the cost, while even Qualcomm and Samsung still license ARM cores.

    • Cl

      I don’t see anyone else coming out with chips for ar/vr

  • Mike

    The section of this article about resolution is a bit misleading, when it talks about 2K and 4K. It can make people think you mean 4K as in 2160p, when you actually mean ~4,000 vertical lines – 4300, which is just short of the vertical resolution of 8K (4320p).

    • Ben Lang

      You’re right, I could be more clear about that.

    • Dave

      Yep I would be one of those folks, I’ve always considered a per eye resolution of 3840 × 2160 to be 4K per eye and something like the HP Reverb G2 2160 x 2160 to be 2K per eye. The Pimax Crystal and Aero to be around 2.5K per eye. Sounds reasonable. I think you are overcomplicating it because you are never going to get a consensous here with so many screen ratio variations, different binocular overlaps, display overscans etc.

      • Mike

        So just don’t say 4K. Say the vertial and/or horizontal resolution. Or say “4K vertical”, etc.

        The detail in the central, main part of your vision is the best reference frame for comparing to TV resolution. That’s determined by vertical lines of resolution. Horizontal resolution has various degrees of overlap between eyes, so it’s not a consistent metric.

  • I bet they are going to compete with Apple. Google is not joining the race to compete with Meta, they want to compete with Apple again. Plus all the rumors talk about Samsung working on a high-end headset

    • Christian Schildwaechter

      It makes almost no sense for anybody to try to compete with a company that sells their product at cost or subsidized, eats billions in development costs each year, doesn’t expect to make money from it for years or make back the development costs for decades. That pretty much a guaranteed loss.

      The sole reason why Meta even tries to establish XR is to break the money printing mobile duopoly of Apple and Google, and become a third major platform taking a fee from everyone. Which is a very expensive long term bet with lots of potential failure points. So the only reason to try the same would be believing in Meta’s approach, having the same goals and billions to burn, which is why only ByteDance with TikTok money even bothered, and they now also u-turned to copy Apple.

      Meta’s product strategy makes sense only for Meta, so the answer to “Is company X trying to directly compete with the Quest?” will almost universally be “No!“

      • Andrew Jakobs

        Sorry, but it is Apple who copies from everybody, the only ‘original’ idea with the AVP is the crowndial, the rest has been done/shown before by others like Meta.

        • Christian Schildwaechter

          The WOW AVP triggered even in VR veterans was due to its extremely intuitive use through eye tracking combined with hand gestures, plus seamless integration into Apple’s ecosystem and apps. None of that is about hardware, which is becoming commoditized anyway, with mostly variants of Qualcomm’s reference design.

          One can of course argue about the SoC or Fresnel vs. aspheric vs. pancake, but that’s just technical details. Macs used x86 hardware from 2006 to 2020, so technically the whole platform was just a PC copy. But users barely care whether a Mac runs on 68K, PowerPC, x86 or Apple silicon, again just a technicality. The Mac’s real value is the user friendly MacOS (a Unix/NeXTSTEP copy). And what will make or break XR is how useful/usable it is, which is where Apple shines and innovates.

          Samsung isn’t going to copy the expensive AVP hardware, instead stick close to Qualcomm’s reference to reduce costs. But what Samsung and everybody else will copy is Apple’s XR user interface, that some Unity developers had reimplemented on Quest Pro just days after the AVP presentation. The crown dial is nice usability feature, but identifying it as the only “original” idea completely misses that (AVP) hardware is just a means to an end. It is all about UI/UX.

          • XRC

            Whilst training as industrial designer in early to mid 1990’s, we were strongly advised to study Apple hardware and software. Not much has changed!

          • Christian Schildwaechter

            Which is a very sad truth. The fact that MRL burns through billions every year, and we still complain about the Quest UI that has barely changed from the Rift in 2016, is shameful.

            That someone implemented the AVP “look and pinch” UI within days not only for Quest Pro with eye and hand tracking, but also with just head tracking and button press on Quest 2, which works much better than Meta’s laser pointer UI, is simply embarrassing.

          • Andrew Jakobs

            But even the UX isn’t designed by Apple, it’s taken from many examples way before the presented it.

          • Christian Schildwaechter

            Again, “XYZ did it first” is silly, because that’s often an unusable a demo or proof of concept. You’d reduce all of VR to Sutherland’s 1968 “Sword of Damocles” with headtracked stereoscopic 3K displays. Only in b/w, suspended from the ceiling and unable to decently render a wireframe cube, but “technically” everything after that is a copy. And SoD a copy of Wheatstone’s 1838 stereoscope.

            Carmack showing Luckey’s duct-taped DK1 prototype at E3 2013 still triggered modern VR, even though Luckey copied existing HMDs, used magnifying glasses, a cheap display and took ideas from MTBS3D and Bolas’ USC Mixed Reality Lab, and Carmack just improved latency on the cheap IMU. Pre-distorting shaders already existed, so by your definition, they did nothing new.

            They still deserve a lot of credit for everything we have today, because they packaged it into something actually useful. Like Apple with AVP, which required lots of work/research, and according to those who tried it, isn’t just a Quest Pro with Apple logo at double the launch price. Apple hired a “Senior Display System Engineer” in 2015-02, with one requirement being “… must understand the key issues associated with developing extremely high fidelity VR environments.” Very few people qualified for that a year before CV1/Vive launched.

      • Dragon Marble

        What we’ve learned from the crash of the early hypes is that XR has to be a long-term investment. Any company aiming to make a quick profit now is doomed to fail. I have said the following every time that happens (Pico, MS WMR): When the dust of XR history settles true visionaries will stand out from all the opportunists. Sony might be the next if they are unwilling to cut PSVR2 price and take some losses. Meta’s strategy makes sense for VR in today’s form factor.

        • Christian Schildwaechter

          Name one company that developed VR products in the first huge early 1990s VR rush with lots of venture capital flowing, that is still alive and working in VR (I can). There is long-time investment, and there is “the tech isn’t ready yet”. Everybody investing into manufacturing holographic consumer displays today is not a visionary, but a lunatic.

          Sure, if you have billions to spare, you can buy twitter and drop its value by 75% in a year, or pay-to-win VR into the market to not pay 30% to Apple and Google. But not wanting/being able to spend billions on consumer VR they might never get back doesn’t make companies opportunists just looking for a quick buck. It mostly makes them responsible businesses.

          • Dragon Marble

            Wait, how does anything go from not ready to ready without any “lunatic” investing in it? VR headsets don’t grow from trees. Give Mark Zuckerberg some credit when it’s due. He seems to be the only one in the industry not surprised that VR is not money-making machine yet.

          • Christian Schildwaechter

            You get there by keeping it in lab until you have a workable prototype. If it’s usable with training, you sell it to industry/military with high margins while slowly improving it, then expand into more niches. Once the prices are bearable, you add prosumers and enthusiasts. Only after hardware is cheap to produce and idiot proof, you consider the low margin consumer market. Holographic displays work as tiny lab samples requiring a room full of lasers. Everybody investing into manufacturing them for consumers is decades early and a lunatic.

            VR didn’t die in the 90s, the military used expensive HMDs and universities/industrie used backprojection VR caves. Oculus aimed to create a gaming PC VR HMD for enthusiasts capable of dealing with issues. Had they stuck to that and released Rift 2/3…, VR tech would be much further, they’d saved a lot of money, but the market would still be just enthusiasts plus professional VR currently served by HTC etc.

            Switching to mobile VR for the mass market to beat Apple and Google came from Facebook. The technology wasn’t ready, and with Quest selling nowhere near mass market numbers, and getting abandonned by 2/3rd of the users, it seems still not ready for mainstream acceptance. And very few in the industry are surprised by that.

          • Dragon Marble

            The problem with your argument about sticking to PC VR is that you are thinking only about hardware. No one would be able to develop VR software and make money if Meta didn’t sell so many headsets. Can we just accept that as a fact because developers told us time and time again.

            If Pico and MS weren’t surprised by the market reception, why did they enter initially and then quit soon after. Only Meta had a long-term mindset at the get-go. Even Sony thought they could take a ride after Meta made VR much more popular, and later was obviously surprised by PSVR2’s low demand. Valve also apparently scrapped the other two planned VR titles after seeing Alyx’s performance. They all seemed to have expected quick money, and pivoted as soon as they saw some red. Only Zuckerberg is saying: we are losing tens of billions a year, and that’s exactly according to plan.

            Not only that’s not irresponsible, it’s the responsibility of giant companies like Meta to bet on big things. It’s either them or the military — the latter generally is more about destroying things than building them. It’s actually quite irresponsible to sit on a money printer while investing and innovating nothing — which is what Valve is doing.

            Success is not guaranteed — we call it a bet for a reason. However, personally, the past few years of my life would’ve had much less joy if it weren’t for Meta.

          • Christian Schildwaechter

            The first games played with DK1 and DK2 where regular PC games with stereoscopy and headtracking hacked in, and early on games like Alien Isolation or Eurotruck Simulator integrated VR modes. Basically what we are seeing now with PSVR2 hybrid games or VR mods on PC, both a much more sustainable model, because these games can distribute their development cost over 100% of the market instead of less than 5%.

            Pico for years mostly released headsets targeting businesses, simply because they couldn’t afford to launch a large consumer release, let alone expand worldwide. Bytedance changed the course after they bought Pico and quickly released the Pico 4 substituted with TikTok money. They were probably surprised with the very limited response, but they weren’t really a VR company, instead a social network that got very rich and decided the best way to proceed was to copy Meta. And Sony’s relation to PSVR2 is very ambivalent, with Playstation CEO Jim Ryan being known to not be happy at all to have inherited PSVR from his predecessors.

            I’m not going to speculate about anything that Valve does, because even Valve usually doesn’t seem to know what they will do in the future and projects rise and fall with how many of their employees support them. But they are most certainly not looking to make a quick buck. Maybe in Valve time, but that’s not quick by anybody else’s definition. And pretty much anybody expected VR tech to develop faster, as seen in early estimates by Michael Abrash, head of VR research first at Valve and then at Oculus/Meta.

            And no, Zuckerberg is not a visionary that alone saw that VR would take decades to mature. That view is twisting reality. The development style at MRL in fact shows a discouraging lack of vision, throwing things at the wall to see what sticks, completely misjudging the market for Quest Pro, trying to sell Mixed Reality as the best thing since sliced bread, with a create/cancel cycle for miserable social VR spaces, where their core competency should lie.

            Zuckerberg is a power monger that controls a lot of money and is willing to use that to gain even more control, even if it takes much longer than he and his VR team expected, because VR turns out to be a lot less useful and attractive to most people than hoped. He tried to break the Apple/Google duopoly before, by offering free mobile data services in developing countries, where the “free” would be limited to a couple of sites, incl. Wikipedia, but of course also Facebook. The Indian government saw this as the market manipulation it was, so they passed laws that “free” services were not allowed to prefer specific sites. No longer able to distribute Facebook phones that channeled users only towards their own offers, Facebook dropped the whole program they had claimed was introduce to close the digital device, after having spent billions.

            Meta is very rich and power hungry, and that’s why doing something that would be insane for any other company makes sense for them to get their own platform with billions of users. But that doesn’t mean at all that anybody should follow their example, or that the market will benefit from it in the long run. Sure, it’s nice we get cheap HMDs, but as nobody can make any money in consumer VR with Meta selling at production costs and eating billions in research, competition has vanished, while there were several PCVR HMD manufacturers. Lack of competition leads to less innovation, and everybody is now hoping that Apple with equally deep pockets will shake things up at the high end.

          • Dragon Marble

            You can call it “throwing things at the wall”, but all innovation requires experimentation. We can all play Sunday morning quarterback, but the valuable lessons Meta learned by spending their resources are benefits to all.

            “Meta is very rich and power hungry” — and very successful. If you scoop up all the bad things about capitalism and throw it at one name, it only evidences your own bias against that entity. Also, just because monopoly impedes competition, doesn’t mean it impedes innovation. Anyone who has promising ideas can simply knock on Meta’s doors. In many cases concentration of resources promotes innovation. Without that scale of investments, we wouldn’t have so many satellites circling Earth today.

            Having said that, Meta is not a monopoly. Small players can’t compete with Meta, but the big players can. They just either don’t have the vision, or have too many short-term investors to answer to — those who want nothing but quick bucks.

            Also, what’s the point of competition if there is nothing to compete for? VR market would’ve been dead if it weren’t for Meta. Your imagined small, niche, enthusiast market cannot exist because VR headsets are not like antique cars. It requires software to function, which can only be developed if there is a large market to share the cost.

  • ArtB

    This article is out to lunch or written by a hopeful apple fanboy. Clearly Samsung is going all in and will launch their own 4k OLED display with superior pancake lenses.

    • Dave

      That’s just rediculous. Put it this way, I can bet money on this headset costing less that 3000 dollars. The XR2 architexture doesn’t have the polish it needs to create a refined XR experience. It will be much improved on the Quest 3 yes, but without more dedicated visual processing power, it will fall behind the Vision Pro, and that’s ok, it will be a cheaper product. Ps I’m no Apple fanboy either.

      • ArtB

        Says the Apple fan boy lol

    • ViRGiN

      Yeah man, and it will be powered by BIxby, the superior AI assistant!

      • ArtB

        By Google assistant and or Chat GPT, both of which really are superior by miles. But don’t worry, Apple will sell dozens of vision pros lol.

        • ViRGiN

          You are so obsessed with potential Apple sales numbers.
          It’s not really a consumer product lol.

  • Christian Schildwaechter

    “Standardization” goes even further. Qualcomm designs the HMD, but all their reference HMDs are build by Goertek, who also produce all Quest, Pico, HTC standalone headsets, and pretty much everything else VR. Goertek does XR research and offered pancake lenses before Meta. So not only is everybody using a similar blueprint and components, but they were created with the party that will most likely produce the headsets. Qualcomm/Goertek provide a XR HMD boutique with everything from templates to components and the complete software stack, allowing companies to quickly create their own HMDs and focus on some specific details.

    Which is a good thing that significantly reduces development costs in the tiny XR market, just like reusing phone technology made VR feasible. This isn’t VR specific, you can create your own mobile phone brand on the website of an OEM handling everything from building to payment and shipping to customers.

    A recent YouTube deep dive into the Steamdeck APU by “High Yield” revealed about 1/8th of the silicon being covered by unused computer vision cores. It turned out that the Magic Leap 2 uses the same APU, with Valve sort of reusing the AMD APU to limit cost/risk on the yet untested Steamdeck. The CV cores (with BIOS settings) triggered speculation about Valve’s Deckard, but the recent OLED Steamdeck comes with an upgraded APU without the CV cores. The APU was actually developed for an unreleased Microsoft Surface, then reconfigured for Magic Leap, then reused for Steamdeck, with Valve hinting that Deckard will use a successor. A lot of cool tech made possible by standardized/reused components bringing down cost.

  • Supeman

    Have a Samsung Gear VR 2 headset sitting around still, abandoned since Android 12. No chance of falling for Samsung’s pump and dump scheme again. Bought the hardware, bought the apps, then Samsung abandoned it. Now they quasi want back in just because they’re afraid of Apple? They should have been more afraid of the vr/ar adopters they already chose to burn.

    • ViRGiN

      It’s your fault you got very invested into an experimental platform that was supposed to open the doors and show the path forward.

      Every company has duds. Nintendo has WiiU, Sony has PSP Vita, Valve has Steam Machines and Index, Facebook had Facebook Phone etc.
      Maybe you should stick to kickstarters? You know, companies with zero backgrounds might satisfy you more.

    • Andrew Jakobs

      But it was more Google’s fault for dumping Daydream. Samsung had two fine WMR headsets (ok, it’s still just crappy WMR). Samsung has been working on VR for years, this got nothing to do with Apple.

  • Andrew Jakobs

    Uhh, it’s disgustingly shameless how Apple openly copy from other companies with ZERO sense of professionalism. But Apple fanboys and girls always think the sun shines out if Apple’s ass.

    • kool

      Apples true innovations are in consumerism

  • psuedonymous

    It’s worth remembering that Qualcomm release a reference HMD design with every XR chip release. Thus far, no consumer devices based on the reference design have been released. All XR-chip-using HMDs have been internal designs into which the XR SoC has been incorporated, not reference designs modified for production. The reference HMDs have proved poor predictors of actual device specifications or designs.

    • MackRogers

      Great point, Qualcomm has been smoke and mirrors and the press has been complicit in the stock pump.

      Qualcomm has an army of patent lawyers which has suffocated the industry for over 2 decades, and have slow-rolled innovation with their 10% performance “improvements”. Great business model honestly if you are a horrible corporate entity, I can not deny that.

    • Ben Lang

      How do you know no headsets have boards or components based on the reference design?