Intel says ‘Project Alloy’ is an open reference platform for companies who want to build advanced, all-in-one VR headsets. But the headset itself still in ong going development. Developers working closely with the headset have shared new details on what’s coming in the next version.

Calling Project Alloy just a VR headset is a bit of a misnomer. Not only is it a sensor-laden device capable of understanding the environment around it, it’s actually a full blown x86 head-worn computer running a PC-class Intel processor. You can literally plug it into a monitor and you’ll see a Windows 10 desktop.

Intel says that Project Alloy is an open reference platform; they want to let companies interested in building similar products use it as a starting point for their own headsets (and of course Intel then wants to sell those companies the chips to power those products). But Project Alloy itself is still in ongoing development, and isn’t expected to reach a level of finality until later this year.

SEE ALSO
Hands-on: Intel's Project Alloy 'Merged Reality' Roomscale Multiplayer Demo

While some aspects of the headset have impressed us (like the field of view, inside-out tracking, and is all-in-one nature), there’s still much room for improvement. A new version of the headset is on the way, and developers working closely with the device have filled us in on some of the forthcoming improvements.

Computer Vision Upgrades

Photo courtesy Intel
Photo courtesy Intel

Perhaps most importantly, one of the major changes coming to Project Alloy is significantly enhanced computer-vision capabilities. Today’s Alloy prototype uses Intel’s RealSense camera technology for real-time depth mapping, but the high latency, low field of view, and sparse density of that depth data leaves a lot to be desired.

The next Alloy prototype will include a next-generation RealSense camera which will be paired with the dedicated Movidious computer vision processing chip, tech which intel acquired along with the company last year. The combination of these two things should mean much better and much faster depth mapping, which Intel is going to need to pull off its vision of ‘merged reality’, which aims to bring parts of the real world into the virtual environment.

Improved Field of View

intel project alloy (3)Another incoming change that surprised us is a supposed increase in the headset’s field of view. Intel hasn’t detailed many of the basic specs of the current Project Alloy prototype, including resolution and field of view, but in our hands-on time we were quite pleased with the FoV, which felt comparable to the Vive and Rift. Like those two headsets, the present Alloy prototype is already using Fresnel lenses, and it’s not quite clear exactly what will be done to improve the field of view further, but we certainly welcome any enhancement on that front.

More Power Efficient CPU With Enhanced Integrated Graphics

intel-7th-gen-core-processor
Intel’s 7-gen Core processor | Photo courtesy Intel

The next Project Alloy headset will also come equipped with Intel’s latest 7th-gen ‘Kaby Lake’ processors. The company says the new chips are faster and more power efficient than their predecessors, which will be important for the standalone headset’s battery-powered runtime. The chip also has improved on-board graphics performance which is also hugely important as the headset doesn’t have the benefit of hooking up to a powerful desktop GPU.

Project Alloy is already a good start for Intel and the improvements coming in the next prototype promise to make it ever better. Another area we’re hoping to see the company make progress is on the bulk of the headset. Yes, for a completely self-contained device that’s crammed full of 10 individual sensors, it’s pretty compact. But it still needs to get smaller and lighter to become a headset that’s practical for regular long-term use.

SEE ALSO
High Res Intel Project Alloy Photos Glimpse IPD Adjustment, Sensors, and More

Presently, it isn’t clear how much Intel might be able to shrink the device. Physical limitations of lens-focusing capabilities mean headset’s like Project Alloy will always need a certain amount of bulk simply to get the lenses and display positioned correctly, but smarter integration of the onboard electronics and sensors into the head-mount is probably a good next step until advanced display technology can solve the focal-distance issue.

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 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."
  • Xron

    So it will be like upgraded version of gear vr? only this one will be powered by intel processor instead of phone? Are you kidding me…?
    We can’t even get decent perfomance while playing games that look ~5 years old, using gtx 970 ( far far superior card than intel integrated card). Its like ps4 (not pro version) without cables.
    So we need to wait for whole year to get ps4 vr graphics from intel?… this year feels dim for vr. Because I think oculus and vive will gets us new hmd’s only by 2018…
    They won’t make them earlier, because there won’t be enough users to power their new hmd’s, until next video card hatch.

    • VR Geek

      The graphics are not the selling point here, it is inside out tracking that is. Graphics are way less important at this early phase of VR. Some of my most favorite ViVe experiences have graphics that even a smart phone could do today like Vanishing Realms. Even if they had to dumb them down more to keep 90fps, that would not matter if the device is room scale, has motion controllers, does not require a PC and/or GPU and has no wires or external trackers. Whoever can deliver this experience first at the right price point is going to appeal to a large audience. Intel is 100% on the right path here.

      • Ian Shook

        I agree. Gamers born before 2000 know what games used to look like and we loved them anyway. Graphics aren’t the answer. Ease of use and just doing it right is. Graphics come later.

        • JustNiz

          Of course gameplay is important to enjoying a game, but if you’ve got nothing good to look at, why even play it in VR? The WHOLE POINT of a VR headset is to have something good to look at!

          • pineanas

            The ultimate point of VR is to be fully immersed and transported to the game world. To transport your body into another reality. Not to look at something nice. Flawless Inside out, room scale and body presence in my opinion are the biggest deal. Otherwise its just a nice display taped to your face:)

          • Ian Shook

            I’ve played Fantastic contraption and really liked it. The most annoying thing about the game was knowing I had a cable behind me – and the UI is a little wonky. But the graphics by any measure are shit. It was still fun though regardless of the cartoony graphics. I guess my point is that personally I think it’s okay for a game to ‘look’ sub-par if it functions ‘above par’ and right now hardware (trackers, cables, headphones) are holding back games from being all they can be. Imagine you want to play doodlejump or some other simple game on your phone. But before you can, you have to position some trackers in your room, plug in your phone, and go through a few menu screens. You’d never play. I think something good to look at, definitely has it’s place, but nobody will want to bother if it’s annoying to ‘get there’. I think I’m rambling. Sorry.

      • user

        does alloy have motion controllers?

        • benz145

          None announced yet, though they have showed off the headset using some third-party controllers.

          • user

            in the first presentation they tracked the hands with cameras on the headset but the fov was very small and i dont see how it could be big enough. at ces they had a presentation with an untracked controller. i wonder where theyre going with it.

      • JustNiz

        > Graphics are way less important at this early phase of VR.
        Speak for yourself.

        > even a smart phone could do today like Vanishing Realms.
        No it really couldn’t, not at 90hz and without massive latency.

        > Even if they had to dumb them down more to keep 90fps, that would not matter if the device is room scale,
        They would, and yes it would.

      • NooYawker

        The point of VR is being transporter into a digital environment. Low graphics also causes more motion sickness, at least for me it does. So graphics is a vital part of VR.

    • Andrew McEvoy

      True enough. Personally for 2017 I have higher hopes for more Triple A games for the current Gen of headsets than a hardware update. Both CV1 and Vive could do with this and also maybe a price drop down the road.

    • JustNiz

      You can bet that for playing AAA games, it will require streaming the video to the headset from a far more powerful PC. All the headset will do is uncompress and display the video stream being rendered/compressed by your PC’s GPU (or possibly an online streaming gaming service). Hello latency, goodbye image quality.

      • benz145

        I don’t think so. For some VR headsets maybe, but that would defeat the all-in-one nature of the headset, which I think is what Intel is aiming for. Don’t forget though, there’s great VR experiences out there that look great, but don’t need to look photorealistic.

    • VRgameDevGirl

      I’m guessing you haven’t played Arizona sunshine. It’s pretty much AAA. Looks fantastic. Runs 90 fps on my gtx 970….. on both rift and vive. and it’s an indie game…

      • Xron

        The problem is that gtx 970 is far superior than whatever they will implant in their “All in one headset” and remember, games that will require more processing power might be coming out soon.
        I can’t understand what do they want to achieve while releasing a headset like this, with outdated capabilities… (Just like switch, though switch atleast has some killer apps from the start),

  • JustNiz

    Fresnel lenses have significant drawbacks in visual quality such as banding, narrow sweet-spot, god-rays etc, yet all the current crop of headsets are using them. The last prototype version of the Rift apparently used conventional lenses and by all accounts gave a significantly better visual experience because of it. It seems the market opportunity for a replacement (non-fresnel) lens kit for the Rift/Vive is probably massive, yet still no such kits exist.
    Can anyone more knowledgeable than me confirm that Fresnel lenses are only being used because they’re cheaper/lighter and that there’s no actual technical reason why non-fresnel lenses couldn’t be used, as long as you were prepared to put up with the extra weight/cost to get a significantly better image?

    • user

      there were rumours that apple works with zeiss. i wouldnt be surprised if they work on lenses for vr

    • Justos

      Fresnel lenses have a larger sweet spot. Oculus’ hybrid lenses are the most clear across the lens as well.

    • benz145

      I don’t think Oculus or HTC/Valve would have sacrificed the visual degradation if it was only for the cost difference. It’s possible that it was for both the ‘sweet-spot’ as Paulo mentioned but also to make the display assembly more compact by allowing for a lens with nearer focal distance (meaning the lenses could be closer to the screen while maintaining the correct focus), allowing the entire headset to be a little more compact. That’s just a guess though.

    • Da Mo (JFlash)

      Fresnel lenses have a larger sweet spot, but i preferred my DK2 to my Vive for the larger FOV seems the closer the Fresnel lenses get to my eye the blurrier the picture so the less you can see to the sides. The God rays are relay bad in light on dark like space sims but unnoticeable in brighter scenes e.g. The Lab. You can’t change the lenses on the Vive well not with a lot of effort check out the tear-down.

      • Da Mo (JFlash)

        ill take that back seem you could get the lenses out with a sucker though normal? lenses would diffract the light and the software has no compensation

    • KOHS

      Fresnel lenses are most of all used for their high magnification with less spheric aberrations then conventional lenses thanks to the possibillity to combine aspheric profile within a low thikness

  • Technically speaking is more than virtual reality because it’s “augmented virtuality”. Anyway, I’m very interested by this, because it is the headset that started the standalone revolution

    • OgreTactics

      It’s not it is in fact, true “virtual” reality not half-assed “VR” or “AR”

      • What do you mean by “half-assed VR or AR”?

        • OgreTactics

          Well, I don’t think there is (or should be) any difference between “AR” and “VR” or “MR” despite bullshit buzzword because they’re all part of the same virtual spectrum, or as you said “virtuality”.

          The fact that we differentiate the two because we’ve been sold that VR headset should or would only be occluded (enclosed) VR headset on one side of the market, and that there are supposedly AR glasses on the other side, is a conceptual non-sense: -Virtual- without AR, meaning your environment is occluded, not tracked, not interactional, not contextual is half-assed VR, as for AR glasses they’re not even real product because the underlaying “glass” technology is 15 years out in the future.

          In reality, we don’t even have one true Virtual Headset yet (which should arrive as soon as we have a Tango Daydream enabled device), and we will do both AR and VR in a -Virtual- Headset form-factor for years until technology has evolved far enough to that we start switching to -Virtual- glasses form-factor that will do both too.

          • Well, actually scientists have given these names: there is a full spectrum between Virtual and real and different parts of the spectrum have different names. Quake played on a PC is anyway virtual reality since you play in a completely virtual world.
            I agree with your vision of everything, but about names I don’t, since I’ve studied them and they aren’t only buzzwords (even if marketing people use them as if they were)

          • OgreTactics

            Well maybe they are indeed “marketing” words. Or rather conceptual names rather than mechanical, technological or scientific ones. In fact people don’t even agree on terms like MR and AR. Then they have a use in differentiating approaches, conceptualisation or development of specifics apps.

            But I think that in order to do conception or create with a larger image it is better to think from the get-go as the virtual or headset, as a continuum because after all a virtual app should work independantely of wether it is in a virtual environment or an augmented environment, it should adapt and let the underlaying tech do it’s job, which in fact should also do it with the same large image in sight for it to be optimised.

  • OgreTactics

    “doesn’t have the benefit of hooking up to a powerful desktop GPU” bafflingly stupid. This looks like a great headset to experiment with…experiment not use if it’s tethered to an isolate computing core and not even to a smartphone.

  • Steve Dennis

    Wow, those Fresnel ridges though…