Raw Data is an early access game for the HTC Vive, one which requires space to move and which allows the Vive to show off its tracking ability. [H]ard|OCP wanted to see how the GPUs found in most high end systems would perform in this VR game and so grabbed several AMD and NVIDIA cards to test out. Benchmarking VR games is not an easy task, instead of raw performance you need to focus on the dropped frames and unstable fps which result in nausea and a less engrossing VR experience. To that end [H] has played the game numerous times on a variety of GPUs with settings changing throughout to determine the sweet spot for the GPU you are running. VR offers a new gaming experience and new tests need to be developed to demonstrate performance to those interested in jumping into the new market. Check out the full review to see what you think of their methodology as well as the raw performance of the cards.
"Both AMD and NVIDIA have had a lot to say about "VR" for a while now. VR is far from mainstream, but we are now seeing some games that are tremendously compelling to play, putting you in middle of the action. Raw Data is one of those, and it is extremely GPU intensive. How do the newest GPUs stack up in Raw Data?"
Here are some more Graphics Card articles from around the web:
- PowerColor Red Devil Radeon RX 470 @ [H]ard|OCP
- Radeon RX 470 @ The Tech Report
- PowerColor Radeon RX 470 Red Devil Review @HiTech Legion
- Sapphire Nitro+ RX 470 OC @ eTeknix
- The AMD Radeon RX 470 4GB Review @ Hardware Canucks
- AMD RX 470 @ Hardware Heaven
- Sapphire RX 470 Nitro + OC 4GB @ Kitguru
- Asus RX 470 Strix Gaming OC Aura RGB 4GB @ Kitguru
- ASUS Radeon RX 470 STRIX OC 4 GB @ techPowerUp
Well that game doesn’t work
Well that game doesn’t work right on AMD parts. Fury X below 1060…
Thank you for the summary Rem
Thank you for the summary Rem
It would be to the benefit of
It would be to the benefit of PCPer readers if you gave us a quick summary of the results. If you keep pushing all of PCPer readers to other websites to read the whole article they are going to just stop coming to your site and go to the other sites.
Uses EU 4 engine and the
Uses EU 4 engine and the console version of the engine is way ahead of the PC build.
Last build DX12 still in experimental stage after 1yr.
I thought pascal is double
I thought pascal is double the performance in vr compared to maxwell, meanwhile 980ti performed similar to the 1070. Maybe it has to have Ansel. Anyway they recommended the 1080 for a good experience. $1500 for the gpu and vr hardware seems expensive now but should come down in the next few years. If volta rumors are true than we should see a titanx pascal like volta gpu 1170 in the $400 price range in the end of 2017.
“I thought pascal is double
“I thought pascal is double the performance in vr compared to maxwell”
Only when using the Pascal-specific extensions like using SMP for lens-matched shading. Without those implemented, the scaling is similar to raw performance scaling.
Simultaneous Multi
Simultaneous Multi Projection(SMP) was not enabled in the build of the game that they used for these tests, so the advantage that Pascal has over Maxwell was not being utilized.
That’s why I’m sticking with
That’s why I’m sticking with my 980 Ti, good enough for VR. Maybe next year upgrade it if Volta is any better.
Yeah two of two sources say
Yeah two of two sources say Nvidia is better at VR. Steam and now Vive.
Maybe AMD’s aging GCN architecture can’t handle new tech like VR as good as Nvidia can with or without async support. There’s something to be said for new hardware after all instead of rebranding.
What graphics APIs are they
What graphics APIs are they using, as Vulkan/DX12 games are still being optimized to work with the new graphics APIs so really, anyone, Nvidia or AMD fans, would be foolish to declare a winner this early in the game. Most benchmarking software still has not been updated to work with the newer graphics APIs in a ways that make any testing beyond suspect. Then there is the fact that Polaris was never intended to compete GPU to GPU with the GTX 1080/1070 and the RX 480’s/470’s price market is mainstream and not high end as AMD has stated all along.
Even the Vulkan/DX12 graphics APIs are still in their tweaking stages, so if anybody needs a definite winner then ESPN and the Olympics are the place to go for their definite winner fixations, as the API/Games ecosystem is still going to need the better part of a year to get its testing software in order along with the APIs themselves. The real interesting part of the new graphics APIs is the Vulkan/DX12 multi-GPU adaptor managed via the Vulkan/DX12 APIs and that part is still under development for both Vulkan and DX12. The new hardware from both Nvidia and AMD is out there but the benchmarking software and the APIs are still going through their tweaking phases, as well as the time it’s going to take for the gaming titles to be updated/optimized for Vulkan/DX12.
Sure a GPU priced and engineered for the high end market is supposed to beat a card priced at less than half that is engineered for the mainstream market. You, Or anyone else, will not be able to credibly claim a winner for at least a year until the API/games/benchmarking software catches up to the new hardware. There is sill hardware features to get driver support ironed out for, and API’s to be tweaked, and games that target any new features developed, and the VR gaming systems are brand new with their own special sets of problems to be overcome over the next few years.
P.S. just because Nvidia uses
P.S. just because Nvidia uses entirely new names for each of its GPU “Generations” and AMD uses GCN versions does not mean that that AMD’s generations are any less of an improvment than Nvidia’s generations! Really you must look at the hardware features/improvments added and not any marketing naming!
edit: improvment
to:
edit: improvment
to: improvement
You do realize that GCN has
You do realize that GCN has been revised right? Polaris is very different from the GCN parts from 2011.
VR does not have some magical difference that makes it inherently different to render for than high resolution, refresh rate monitors.
I think the correct answer is that that unoptomized DX11 games simply run better on Nvidia hardware. That doesn’t mean that going forward when games on newer APIs are released that AMD won’t be competitive.
WRONG – Polaris and pascal
WRONG – Polaris and pascal are new architectures, neither one is rebranded.
Per that article, the only previous generation card that was tested was a gtx 980ti, still not a rebrand.
So exactly where is this “rebranding” you refer to?
polaris is not really new.
polaris is not really new. just GCN with added some new feature. and die shrink. the difference between polaris and previous gen GCN is not big like how different it was HD6970 architecture compered to HD7970 architecture.
Yes, it really is new. It
Yes, it really is new. It isn’t simply “just GCN with added some new feature. and die shrink.’ it has several new features added along with many changes to get it to the new smaller node. You should read about all the changes that were made instead of posting an opinion as if it were truth.
Yes I realize they do update
Yes I realize they do update it some each generation but the basis is the same. Polaris is still GCN so it is not as energy efficient as it should be on a 14nm node compared to Nvidia on 16nm node. Vega is supposed to be the “new” arch. Oops, it is not it is based on Polaris sorry.
This is how AMD can make one driver that gives performance benefits back to GCN 1.0 even. The newer hardwares gets the biggest benefits as they have more features and or more hardware than older GCN cards. More ACE units, better tesselation support, and etc.
GCN is full of rebrands. For example 7970 became the 280x the 280x became 380x and now Polaris could be 380x node shrunk. They are too close to be coincidence.
http://www.game-debate.com/gpu/index.php?gid=3489&gid2=3130&compare=radeon-rx-480-4gb-vs-radeon-r9-380x.
Maxwell was new arch as is Pascal supposed to be yet it is always referred to as Maxwell on steroids or others.
You putting your foot in your
You putting your foot in your mouth again? The 280x did not become the 380x. These were different archtectures, 280x was indeed the 7970, however the 380x was Tonga. Polaris is not Tonga on a node shrink, again, they are different archtectures. Polaris has many new features and changes to get it to 14nm, please get this through your thick skull.
Pascal is not Maxwell on steroid, although they do share some similarities, just like with polaris, pascal has many new features added along with changes to get it ot the lower node.
Yeah but they got exactly the
Yeah but they got exactly the same performance capability. Too similar to be called different.
http://www.game-debate.com/gpu/index.php?gid=2265&gid2=1860&compare=radeon-r9-380x-vs-radeon-r9-280x#
Apparently others must share the same “misconception”
http://www.tomshardware.com/answers/id-2709859/380x-280x.html
PS I love it when you get all childish and insulting. It’s funny.
Apparently I was pretty close anyway. The 380x is based off the 285. Which is a slight tweak of what the 280. What no way you say.
https://www.reddit.com/r/Amd/comments/3ukuvj/r9_380x_vs_r9_280x/
http://techreport.com/review/26997/amd-radeon-r9-285-graphics-card-reviewed
In the gaming community the
In the gaming community the best minds do not reside! The games companies and gaming developers, driver developers, hardware developers are another story entirely as they have to have brains(script kiddies not included). There are no gaming nerds, as nerds do not have time for much gaming what with all that science and math that is needed to actually qualify as a nerd.
The not so bright gaming community by and large is about the same as the sports community with brain power no so much required. The gaming, GPU, CPU, fanboys are at the bottom of the evolutionary charts for simians and even single cell organisms.
And you’re about as highly
And you’re about as highly evolved as something I flushed away in the toilet about an hour ago. So cheers.