An unnamed and fully unknown AMD graphics card has appeared within the OpenVR benchmark paired up with an unnamed AMD CPU. That thriller is just compounded by the consequence, as collectively they managed to beat each Nvidia RTX 2080 Ti on the desk, together with closely overclocked ones, by a sizeable margin. Is that this the “Huge Navi,” GPU that crew crimson followers have been ready for? It certain looks as if it.
Because the launch of AMD’s RX 5700 graphics playing cards in July 2019, hypothesis about when it could launch a brand new high-end graphics card has gathered traction. There have been rumors of AMD workers testing an alleged “Nvidia Killer” graphics card, and CEO Lisa Su has repeatedly acknowledged that we are able to count on AMD to return to the highest of the graphics card efficiency spectrum. However with no announcement of something near “Huge Navi,” at CES 2020, many suspected we’d be ready a very long time to study extra. With this leaked consequence, maybe not.
Found by Reddit consumer MuchCharles, the mysterious AMD graphics card appeared within the OpenVR benchmark outcomes testing on an HTC Vive headset at 90Hz, with a decision of 1,512 x 1,680. It managed a rating of 103.32. For comparability, essentially the most closely overclocked Nvidia RTX 2080 Ti at these specs managed a rating of simply 88.10.
Though the CPU was listed as unknown by the benchmark, VideoCardz broke down the code to find that it’s probably an AMD Ryzen 7 4800H, paired up with a supremely powerful graphics card. Or two.
There are some caveats right here, corresponding to OpenVR being a really new benchmark and isn’t as extensively used as different testing software program, like UL’s 3DMark, so it most likely doesn’t have the broadest part of programs for comparability. Engineering samples do generally tend to throw testing software program off, too. And there’s additionally a slight likelihood that we’re a super-powerful Nvidia graphics card, however that the benchmark detected the 4800H’s onboard graphics because the GPU and named it as such.
However the creator of the benchmark, definitely thinks the GPU is AMD and attracts another attention-grabbing conclusions.
They make it clear that they don’t know any greater than us, and may’t look into the outcomes any greater than the rating and listed . That stated, they level out that normal 2080 Ti graphics playing cards are likely to get round a rating of 80, and the 88.1 rating of the highest 2080 Ti is a closely overclocked model. That implies that this unnamed Radeon GPU is round 17% sooner than the highest 2080 Ti, and shut 30% sooner than a inventory 2080 Ti. That places it exterior the realms of something we’ve ever seen from AMD or Nvidia.
The OpenVR creator highlights that it may be potential for a closely overclocked, liquid-nitrogen-cooled Titan RTX or 2080 Ti to realize such a rating, however that appears unlikely for the reason that benchmarker was utilizing a normal (albeit engineering pattern) laptop computer CPU. In the event that they had been choosing most potential efficiency, they’d have used a desktop CPU and carried out an analogous overclock on that part too, absolutely?
Contemplating it’s extraordinarily unlikely for somebody to have entry to each an engineering pattern of an unreleased AMD CPU, and an Nvidia GPU on the identical time, it’s possible that what we’re is an engineering pattern of an unannounced AMD graphics card. Which means it very properly may very well be Huge Navi. It may very well be a pair of 5700 XTs in crossfire too… however that appears much less probably, as crossfire is successfully ineffective in trendy gaming.
Whether it is Huge Navi, and AMD has a GPU that may beat a 2080 Ti by virtually a 3rd, that’s big information for crew crimson followers. It won’t be sufficient to take the efficiency crown when Nvidia reveals its Ampere GPUs — which can even be primarily based on 7nm — later this yr. However this could nonetheless be by far essentially the most succesful graphics card AMD has ever launched, and provides players and lovers an choice on the subject of high-end efficiency.
It would even do one thing concerning the loopy costs we’ve seen on the high of the GPU scale for the previous few years.