news with liquidvr amd aims to make virtual reality more fluid
News

With LiquidVR, AMD aims to make virtual reality more fluid

Virtual reality is a hot topic at this year's Game Developers Conference, and AMD has its head in the game, too. Today at the show, the chipmaker announced LiquidVR, a set of tools designed to bring about the holy grail of virtual reality: a "motion-to-photon" latency low enough to make the experience subjectively seamless—an effect VR aficionados call "presence."

AMD says this quest involves optimizations across "the entire processing pipeline," from the GPU to the display hardware on VR headsets. Here are the main features of the LiquidVR 1.0 SDK, in the company's own words:

  • Async Shaders for smooth head-tracking enabling Hardware-Accelerated Time Warp, a technology that uses updated information on a user’s head position after a frame has been rendered and then warps the image to reflect the new viewpoint just before sending it to a VR headset, effectively minimizing latency between when a user turns their head and what appears on screen.
  • Affinity Multi-GPU for scalable rendering, a technology that allows multiple GPUs to work together to improve frame rates in VR applications by allowing them to assign work to run on specific GPUs. Each GPU renders the viewpoint from one eye, and then composites the outputs into a single stereo 3D image. With this technology, multi-GPU configurations become ideal for high performance VR rendering, delivering high frame rates for a smoother experience.
  • Latest data latch for smooth head-tracking, a programming mechanism that helps get head tracking data from the head-mounted display to the GPU as quickly as possible by binding data as close to real-time as possible, practically eliminating any API overhead and removing latency.
  • Direct-to-display for intuitively attaching VR headsets, to deliver a seamless plug-and-play virtual reality experience from an AMD Radeon™ graphics card to a connected VR headset, while enabling features such as booting directly to the display or using extended display features within Windows.

The official LiquidVR page mentions cutting motion-to-photon latency to "less than 10 millisconds." That means delivering a solid 100 FPS to the user's eyes—and it's about in line with the target I heard Oculus quote at AMD's APU13 conference a couple years back. I seem to recall Oculus mentioning tricks like time warping, as well, which it said would enable low latencies without requiring the GPU to sustain triple-digit frame rates.

Developers (and users, too) can sign up to learn more about LiquidVR here. AMD has also posted a LiquidVR video on YouTube, but there's not much in there beside a back-to-the-basics explanation of how VR works.

0 responses to “With LiquidVR, AMD aims to make virtual reality more fluid

  1. No problem. It’s [i<] virtual [/i<] reality. AMD only have trouble with actual reality. [quote<] Maybe they should rename the company A.D.D.[/quote<] I'm sure they will when they can find someone with the attention span required to rework the logo.

  2. It’s pretty much the same feature-set as the Oculus SDK, with some hardware specific features, and intended to support more HMDs. That’s probably what Oculus wanted: their SDK to act as an example for other more general SDKs to be derived from.

  3. [quote=”Article”<]Affinity Multi-GPU for scalable rendering, a technology that allows multiple GPUs to work together to improve frame rates in VR applications by allowing them to assign work to run on specific GPUs. Each GPU renders the viewpoint from one eye, and then composites the outputs into a single stereo 3D image. With this technology, multi-GPU configurations become ideal for high performance VR rendering, delivering high frame rates for a smoother experience.[/quote<] I think someone on one of the podcasts had asked if this was possible and I think the consensus was that is would be a great deal of work and probably wouldn't see the light of day. At the time, I agreed with that consensus. Sounds pretty cool, so I don't mind being wrong.

  4. How about you make me a Tonga card with 4GB of ram at 7Ghz and reduce 3D power draw by 30%.

    Please & thanks

  5. AMD, please focus your resources on delivering quality drivers and competitive core products. Nice-to-haves happen after need-to-haves are secured.

  6. Just my observation. I see you disagree and many are with you. Would you rather cite examples that show otherwise? Because if you can’t present any, I would think people simply don’t like my statements but they’re nonetheless true.

  7. Don’t come in here with your LOGIC!

    Let’s just forget about AMD’s original and successful technologies such as ZeroCore.

    Fanboys can easily find things to dig at, and will happily harp on about their deluded view of who is losing.

  8. Well the VR sli and VR direct stuff still hasn’t been delivered by Nvidia – so LiquidVR is just as much hot air from AMD as VR Direct is from Nvidia. It wasn’t even part of the the GDC presentation so who knows how far off it is.
    Having used it on both Nvidia and AMD I can tell you DK2 is still as janky on either platform.
    There’s no hardware environment set in stone yet (let alone final OR environment) so I think there won’t be anything available to us until they don’t have to make the caveat “subject to change…”.

    To be honest I hope it is hardware manufacturer independant (AMD,Nvidia) as mixed platforms would be a nightmare for devs.

  9. So… If AMD copies someone that makes them a mere, pathetic follower, but conversely, if AMD is copied by someone else it means AMD’s version is surely obsolete and will be “left behind”?

    Isn’t that strange?

    I really wonder how much of AMD’s situation is due to how that company is regarded. What can it do if any innovation from their side is deemed a failure by default? I doubt acting like a leader will help – or be a full solution, at least.

  10. Hmmm and it looks as though it’s a mutation of Mantle, which makes sense. Even if these were originally done by Occulus they’re going to keep tacking things onto it that benefit it and build it out. Cool that this stuff is finally getting a name and a place now instead of just developer tech. Hopefully Oculus and AMD keep pushing things .

  11. I’m kinda thinking the same thing. Whatever others are doing, AMD does too (Freesync, Turbo Core). That makes them a follower. Then when they do come up with something before the others, they either largely ignored, lose steam and become abandoned (3DNow!, most probably FMA4, XOP and F16C too, TruForm, TruAudio, maybe HSA as well but it will probably be relegated to niche applications), or get copied by others and leave the AMD version behind (Mantle to be superseded by DX12 and Vulkan).

    If you’re a leader, you ignore what your smaller competitor does and makes a lot of noise about. Instead, you forge your own path and let them all follow you. You also largely ignore them crashing your parties and their potshots. Taking potshots at your competitor only shows that you’re the follower.

  12. Sounds like that list nVidia had last year of things they were doing to support Oculus Rift. Turned out, that list was the same list that Oculus had requested previously.

    nVidia acted as though they’d done it on their own, but turned out they were just doing what Oculus asked.

    The only way that could have been more groan-inducing is if they had done it a year after their main competition and still called it innovation.

    …Right, AMD? 😉

  13. Soooo, Mantle is gone.
    Freesync is here.
    LiquidVR is soon to come.

    Will any of the three be in existence sixteen months from now? Should I trust that any of these technology sinitiative will have enough staying power that I should buy a product that uses them?

    Hm.

  14. Sounds like a point-for-point implementation of the same things nVidia are planning with VR-Direct, as responses to the things Oculus have been asking after. I’d guess this stuff will end up getting integrated into DirectX/OpenGL at some point, but for now it seems like something they need to provide themselves, and stand to lose out when VR hits retail if they don’t get onboard.
    Of course, neither of them can help but package it up, slap catchy names on it, and market it as a fantastic new feature of their graphics hardware 😛

  15. [sarcasm]AND IT’S OPEN, FREE AND CAN BE USED BY ANYONE!!![/sarcasm]

    Maybe they should rename the company A.D.D.