AMD and NVIDIA GPU Vive VR Performance in Raw Data @ [H]

CheeseCake87 said:
I'm not sure you should of done this review at this time to be honest. Hi, by the way, just joined, great to be here. My opinion; Not that it matters, is that I'm greatful for the work you have done here and yeah... good job; but, to early for this stuff, to be honest. No offence... You have taken an early access game (granted nothing is really out yet) and benched it. Not to mention its a GameWorks title. A closed source addition that none of us are allowed to poke holes in, and with nVidias track record... weeelll, you can kinda see where this is going.

I noticed a comment above that was saying AMD should just die bla bla yada. Lets not get into that much; horrid if you ask me. Fact is, companies have no soul. No feelings. No interest in you. They are only interested in the money. Some donate their work to the open source to progess technology and others lie about specs and pay reviewers gospel funds. It is very much a dog eat dog market (if dogs actually ate dogs)

Fact is folks, speculation is only that; Speculation. This review was done way to early. VR early access game with closed source GameWorks, in a market that is still very young. Not even SMP is out yet, and you done the review anyway lol Tame your fleshy flag poll for VR for now. - As the review stands just now, written in a rather bias view, just saying lol - Not that I care. I recommend parts towards clients set budgets; just last week I was plugging in a STRIX 1080 - but if someone wanted a budget system. It has to be AMD. The customer always comes first, your bias should never get in the way of whats best (just a little lesson for the AMD hater guy xD - LOVE YOU!) - k bye!
Click to expand...

Strange you should join here to comment like this.
I would rather there was objective research done so consumers can quantify the state of VR.
It is absolutely welcome given the complexity of the hardware, drivers and software to implement!

You had the option to walk away and learn nothing.
Instead you joined this forum and insisted we should all know nothing.
Baffling.
 
Kyle, why wouldn't you want to get a bit drunk and try a Vive at 45 fps? It's for science!

Joking aside, that was a good read and very interesting. This is the kind of stuff that I'm going to be keeping a close eye on in the coming months. I recently built a PC around a 1070 for the purpose of eventually getting a VR HMD. I'm hoping they get SMP enabled soon so I can see what kind of performance to expect. I'd hate for the limit to be pushed so soon on the 1070 and I haven't even had a chance to get a proper VR set up going.
 
CheeseCake87 said:
I get what your saying about raw data. It is too early to test though. Really, come on... If someone is buying a $700+ card, they are buying to keep it for a while. An investment. It is a little early to be pushing what is best for VR at this moment.
Click to expand...
This kind of reminds me of early AotS benchmarks. Half-a-year prior to release, 1060 being like 2x as fast as 480 and faster than FuryX - makes no sense without using SMP. With AotS, by release, nVidia closed the performance gap to where it roughly should be. The difference might be that this game might actually being played once it is released.

/yawn, wake me up when the game releases (plus has two months of patches)

Sure thing: AMD sure has some serious work to do with UE4.
 
Well, I do know this. Sitting around waiting for VR to become mainstream is not going to answer even the personal questions I have about it. And I was very thorough in pointing out that Raw Data is an early release title and we should expect performance to get better as it moves closer to release. How much better? Who knows.

I thought it was funny this morning as I did more searches on actual subjective VR gaming data....as I want to see how others are doing it, if there were....



Their thoughts in their VR review was "its good enough." They also do not understand how Adaptive Quality actually works. Furthermore they pushed subjective scoring as being the way to measure VR perf the best. I cannot get onboard with that.
 
ppi said:
This kind of reminds me of early AotS benchmarks. Half-a-year prior to release, 1060 being like 2x as fast as 480 and faster than FuryX - makes no sense without using SMP. With AotS, by release, nVidia closed the performance gap to where it roughly should be. The difference might be that this game might actually being played once it is released.

/yawn, wake me up when the game releases (plus has two months of patches)

Sure thing: AMD sure has some serious work to do with UE4.
Click to expand...
Agreed, and I am surprised nobody has brought this up yet either. We would not use AotS till it was released. Then why not with Raw Data? AotS was not much ever anything more than a benchmark, "no one" plays it. Raw Data already has a big following of gamers playing it. I think that is an important difference to point out on that front.
 
I told you to keep testing whatever you want to test. Test whatever vr games of interest comes out with whatever engine/api/optimizations you want.


But if I see a vr game with a 1060 beating a fury x, "I" will judge that game essentially broken as far as performance goes on amd hardware.

Now most of you will just pile onto amd at this point for not being proactive enough to get both the devs and themselves to optomize performance for their cards. Fine, go ahead. But for me, I am at the point where I see any game not released in dx12/vulkan as a black mark where performance is a premium.


You chose a game for a medium that needs as much performance as possible... and you chose a golf cart api vs a real car because the latter required more initial work. It's not the kind of effort that ought to be rewarded compared to others who do put in the extra work.
 
Remon said:
I was talking about the engine, not the test.
Click to expand...
OK, same difference but anyway. Please suggest me a GPU intensive title to test with that you would consider "not biased."
 
I am not tech savvy by any means. I play games. Help me understand why it is incumbent on video game developers to optimize for a certain card? I know they can do that, and it can be mutually beneficial, but not a requirement. Is it not the responsibility of a video card manufacturer to understand the market and produce hardware that is going to be optimal running whatever is out now? When I read a review I want to know what card is performing the best at each price point in today's games, especially the ones with engines I am likely to see the most often. Good job Kyle.

I do want to say that I don't post here often but I read and trust the reviews here. I take offense at sly insinuations that this site may be in the pockets of certain hardware companies. That is utter bullshit. I am not a fanboy of any hardware company, I am of HARDOCP.

tybert7 said:
I told you to keep testing whatever you want to test. Test whatever vr games of interest comes out with whatever engine/api/optimizations you want.


But if I see a vr game with a 1060 beating a fury x, "I" will judge that game essentially broken as far as performance goes on amd hardware.

Now most of you will just pile onto amd at this point for not being proactive enough to get both the devs and themselves to optomize performance for their cards. Fine, go ahead. But for me, I am at the point where I see any game not released in dx12/vulkan as a black mark where performance is a premium.


You chose a game for a medium that needs as much performance as possible... and you chose a golf cart api vs a real car because the latter required more initial work. It's not the kind of effort that ought to be rewarded compared to others who do put in the extra work.
Click to expand...

You have been asked to name the game that should be tested. When are you going to do that?
 
Nenu said:
Strange you should join here to comment like this.
I would rather there was objective research done so consumers can quantify the state of VR.
It is absolutely welcome given the complexity of the hardware, drivers and software to implement!

You had the option to walk away and learn nothing.
Instead you joined this forum and insisted we should all know nothing.
Baffling.
Click to expand...

Thank you so much for your warm welcome. I really enjoy the sense of community.

I also really enjoy chasing reviews on how half built cars with only two wheels turn corners.
- As I said before, these are enthusiast dollar stats. To early to be addressing or advising the mainstream market on what to buy in my opinion. Reviews like Kyles commonly get echoed in a dumbed down style. We have no idea what each company can do in the long term for VR at this moment.

The data speaks for itself, but advising to buy budget hardware based on these numbers at this early stage is not the way to go. It works both ways, GCN may come out on top, but SMP my take the crown. Either way, advising to buy anything budget VR right now is not a good idea for people looking to keep their hardware for some time.

"My opinion; Not that it matters" - right? Thanks for your welcome once again though. Nice to feel accepted :)
 
Smoked Brisket said:
I do want to say that I don't post here often but I read and trust the reviews here. I take offense at sly insinuations that this site may be in the pockets of certain hardware companies. That is utter bullshit. I am not a fanboy of any hardware company, I am of HARDOCP.
Click to expand...
(y)
 
tybert7 said:
I told you to keep testing whatever you want to test. Test whatever vr games of interest comes out with whatever engine/api/optimizations you want.


But if I see a vr game with a 1060 beating a fury x, "I" will judge that game essentially broken as far as performance goes on amd hardware.

Now most of you will just pile onto amd at this point for not being proactive enough to get both the devs and themselves to optomize performance for their cards. Fine, go ahead. But for me, I am at the point where I see any game not released in dx12/vulkan as a black mark where performance is a premium.


You chose a game for a medium that needs as much performance as possible... and you chose a golf cart api vs a real car because the latter required more initial work. It's not the kind of effort that ought to be rewarded compared to others who do put in the extra work.
Click to expand...
Theres not much mention of AMD apart from people whining.
Yell at AMD, you might get a result.
 
CheeseCake87 said:
I also really enjoy chasing reviews on how half built cars with only two wheels turn corners.
Click to expand...
Except you can't drive a car with two wheels, and I can buy Raw Data and play it right now.

Ninja Edit: And a bit more to your point on this. We stopped using "pre-released" games a long time ago because between that time, and the time they went gold, all sorts of things could change. But back then, you did not have "early release" to the public where you could buy a non-gold game. So I do understand what you are saying fully, but the mechanics of the distribution system have changed a bit and I think HardOCP needs to change along with that.

CheeseCake87 said:
Reviews like Kyles commonly get echoed in a dumbed down style.
Click to expand...
I am not going to disagree with you on that one, but I don't think I can be held responsible for that in any way either.

CheeseCake87 said:
As I said before, these are enthusiast dollar stats. To early to be addressing or advising the mainstream market on what to buy in my opinion.
Click to expand...
You are 100% correct here, but HardOCP guys are early adopters and I think they want to know what their money buys them today, not next year. And we have always pretty much stuck with that mantra.

CheeseCake87 said:
Nice to feel accepted
Click to expand...
You are more than welcome here CheeseCake87 , I have no problem with criticism or different points of view. It is quite often the feedback that I get in these forums that shapes the way we create and represent content. Everyone gets a voice here. (Just don't start calling names. That will get you (or anyone else for that matter) the boot quickly. )
 
CheeseCake87 said:
Thank you so much for your warm welcome. I really enjoy the sense of community.
Click to expand...
When you come in guns blazing, dont expect people to think you are here for the community.
We know what its about here.
If you have a point to make use tact.

I also really enjoy chasing reviews on how half built cars with only two wheels turn corners.
- As I said before, these are enthusiast dollar stats. To early to be addressing or advising the mainstream market on what to buy in my opinion. Reviews like Kyles commonly get echoed in a dumbed down style. We have no idea what each company can do in the long term for VR at this moment.
Click to expand...
Its good to know the state of play for both sides so they get a kick up the arse when needed.

The data speaks for itself, but advising to buy budget hardware based on these numbers at this early stage is not the way to go. It works both ways, GCN may come out on top, but SMP my take the crown. Either way, advising to buy anything budget VR right now is not a good idea for people looking to keep their hardware for some time.
Click to expand...
Making things up wont help you.
As you said the data speaks for itself.

I didnt see any recommendations being made.
The Rift cannot play this game, no fault of the reviewer. Oculus might get a move on releasing kit that will allow this now it is exposed. Its their business that needs to survive.
Regarding video card hardware, AMD pushed their cards as VR worthy yet they have shown no advantage despite NVidias features not being live yet.

"My opinion; Not that it matters" - right? Thanks for your welcome once again though. Nice to feel accepted :)
Click to expand...
You started lol.
 
AceCR42 said:
Well, about that...

Ashes of the Singularity: Ashes of the Singularity · AppID: 228880
Players online
  • 91 right now
  • 120 24h peak
  • 739 all-time peak

Raw Data: Raw Data · AppID: 436320
Players online
  • 78 right now
  • 124 24h peak
  • 429 all-time peak

It might get more popular as it approaches full release, but currently seems like player numbers are roughly the same.
Click to expand...

Totally fair argument. How many people have gaming PCs that can run AotS? How many people have HTC Vive that can run Raw Data? Now let's figure out some usage percentages. :)

But back to the point I made. People were not, and still not are by any large percentage PLAYING AotS. It has always been a "benchmark." Surely you can look at it differently if you choose. But I have stated my logic on this and you can accept it or not.
 
tybert7 said:
I don't have one. I'm not jumping into the vr gaming segment at this point so have not paid much attention on the relative performance of different titles.
Click to expand...
OK, so your point is don't show me anything until you are ready to buy into VR?
 
rinaldo00 said:
What new games did you buy? I am always looking for good Vive games to try.
Click to expand...
This is what I have been going through currently.

upload_2016-8-1_18-8-29.png
 
tybert7 said:
But if I see a vr game with a 1060 beating a fury x, "I" will judge that game essentially broken as far as performance goes on amd hardware.

Now most of you will just pile onto amd at this point for not being proactive enough to get both the devs and themselves to optomize performance for their cards. Fine, go ahead. But for me, I am at the point where I see any game not released in dx12/vulkan as a black mark where performance is a premium.
Click to expand...

Man you really do some heavy justifying, but you have one valid point. AMD should do more optimize their drivers for engines, in particular the heavily used UE4 engine.

Also waiting on hypothetical titles that support Vulkan or DX12 is completely irrelevant. AMD either works today, or not, as we can see it's a firm not atm.

Doom is not a VR title.

Your basically advocating buying a gpu that only works on some titles, and refusing to acknowledge the actual market place.
 
tybert7 said:
I don't have one. I'm not jumping into the vr gaming segment at this point so have not paid much attention on the relative performance of different titles.

Also, to clarify with the "you chose" I'm talking about the devs who push on with new projects with lower performance apis. Right now we can still sort of give people a pass as many started on games well before things were fleshed out. But if someone started recently and went with something for a performance constrained game and refused to build it in a way that allowed it to run anywhere near as well as it could, their game should not be looked upon as well as others that built better.

Put another way, the devs that build their games to run like doom should get more attention and praise compared the ones that make ac unity, or the latest batman.
Click to expand...
So you don't have a game that you think would be suitable, but you claim bias because Kyle tested a game that uses one of the most commonly used engines on the market today? Do you see how that might cause people here to take issue with what you've said?

The simple fact is that no matter what game or games that Kyle chooses, he's going to get people who claim bias because the results don't paint their preferred company in a good light. Speaking personally, my issue isn't with the 480 or even with AMD themselves - we all need AMD to do well or we'll end up with $4,000 cards that have no innovation. My issue is with the way AMD markets themselves and their products. More often than not their marketing is the polar opposite of what their products can do in the real world. And THAT is what I take issue with.
 
Aireoth said:
What are your current favorites?
Click to expand...
HordeZ
Raw Data
Hover Junkerz
Out of Ammo

HordeZ and Raw Data will get you to a point that you have to stop you get some damn wrapped up in it. The level of immersion is so incredible that I just cannot put it into words to do it justice.

Some of the others on the list I was supposed to start testing today that I bought this weekend during the Steam sale, so I have not queued up Battle Dome, BellyBots, Space Pirate Trainer, or Zombie Training Sim. (Had to do news today and tomorrow while Steve takes a couple of personal days.) And keep in mind that I am very much a first person shooter guy as well. VR Fun House is very much worth the download too for a demo to see that much physics in action.
 
I want to leave to very GPU-intensive titles, so trust me I am keeping my eyes open around here for feedback.
 
I'm curious. Missing from the card specs is the VRAM size. This should only matter for the RX 480, but if the card used was a 4 Gig card, then both poor performing cards are 4 Gig cards. It would be interesting to see what happens with an R9 390, or even an 8 Gig 480.
 
eachus said:
I'm curious. Missing from the card specs is the VRAM size. This should only matter for the RX 480, but if the card used was a 4 Gig card, then both poor performing cards are 4 Gig cards. It would be interesting to see what happens with an R9 390, or even an 8 Gig 480.
Click to expand...
My bad, it is an 8GB card. I will update. Thanks for pointing that out.

That is fixed now, you may have to reload/F5 the page.
 
For VR performance, Raw Data is a pretty good choice as to a current VR title that can be dialed up to stress a GPU. Elite Dangerous is another - glad you are considering it as well!

Can't wait till you guys get your mitts on the new Nvidia Titan X and put it through its paces, both in 4k and VR.

I really hope the newer VR games being worked right now will start taking advantage of the "new and improved" VR features the Pascal cards and their drivers bring to the table.... guess the game engines adopting them will also help. Hopefully by this winter...
 
Kyle, did you experience theBlu? I'm asking because I did not see it on the list.
It's not a game, but it has great graphics quality. Definitely recommended
theBlu on Steam
 
Nice writeup, and you appear to be the first out of the gate for VR benching! Congrats.
 
Simplex said:
Kyle, did you experience theBlu? I'm asking because I did not see it on the list.
It's not a game, but it has great graphics quality. Definitely recommended
theBlu on Steam
Click to expand...

Awesome game but I'm not sure it's really that graphically intensive. You should still try it though, Kyle.
 
Its visually impressive, which is why I recommend it.
It can be made gpu taxing by cranking up supersampling (renderTargetMultiplier).

Burticus said:
Nice writeup, and you appear to be the first out of the gate for VR benching! Congrats.
Click to expand...
Tomshardware did some VR benching when they were reviewing Oculus and Vive in April.
 
Kyle, any data on if using a 1080 and another older card (980 or 980 ti) will kick in their PhysX offloading in VR funhouse and how much of a difference it makes? Odd that they offload physx and only go SLI with three cards.

Good first pass at the VR review format, whole new ballgame to get used to isn't it? Hope we get some more mature VR games soon, my guess it we'll see a bunch more good ones when the Oculus Touch controllers hit.
 
Discounting all games DX11 because AMD is underperforming even though we all know AMD drivers are shit in DX11. That's real unbiased, objective view of the world right there.

News flash: Most games going forward will still be DX11 for the foreseeable future.
 
Raw data is making me contemplate buying a Vive. Would give my 1070 a nice workout.

As a single guy living alone and with his own house, I think having a dedicated VR portion would be great. 1600 square feet for one person is so much space to me.
 
Simplex said:
Tomshardware did some VR benching when they were reviewing Oculus and Vive in April.
Click to expand...
Please link on that because I find zero benchmarks or GPU comparisons.

TroubleMagnet said:
Kyle, any data on if using a 1080 and another older card (980 or 980 ti) will kick in their PhysX offloading in VR funhouse and how much of a difference it makes? Odd that they offload physx and only go SLI with three cards.
Click to expand...
I think it is currently still restrained because it is SLI, not mGPU/EMA.

noko said:
Also how does one move around in this title etc. I know not a full game review or quality setting discussion. Good review on introducing VR datasets and to allow us to give feedback for refinement.
Click to expand...
It is teleportation. More feedback the better.

collegeboy69us said:
As a single guy living alone and with his own house, I think having a dedicated VR portion would be great.
Click to expand...
Do it. I think it is worth the money now if you are a "big" gamer. Still a lot of titles out there to play with more releasing every week.
 
Simplex said:
You're right, there are no benchmarks. My mistake. I confused Tom with PCPer: Measuring VR Performance and Experiences - A Prologue | PC Perspective
Click to expand...
That was PCPer.com. But they were simply evaluating the displays. They did not performance comparisons on GPUs.

Wow, you Ninja edited me! Good job. ;)

And for what it is worth, I think they started over thinking that a bit. I don't think there is anything to "compare" right now in terms of headsets. There are a 1000 articles about Vive vs Oculus already, we don't really need to discuss that. GPUs need to be looked into now.
 
Sorry, for ninja editing you :) I suddenly remembered the site and hoped noone will notice the edit.
PCPer did do some comparisons of The Gallery, for example:
http://www.pcper.com/files/review/2016-04-05/gtx980tiv970-high.png

I replied to "you appear to be the first out of the gate for VR benching!" by pointing out other sites did VR benching before (not saying they did it better).
Here's another VR bench, comparing 1080 with 980Ti in Oculus games: NVIDIA GTX 1080 Performance Review: Head to Head Against the 980 Ti - Page 3 of 3 - Road to VR
But [H] may be first to bench Raw Data, that I'm not disputing.
 
Last edited:
Simplex said:
Sorry, for ninja editing you :) I suddenly remembered the site and hoped noone will notice the edit.
PCPer did do some comparisons in the gallery.
I replied to "you appear to be the first out of the gate for VR benching!" by pointing out other sites did VR benching before (not saying they did it better)
Here's another VR bench, comparing 1080 with 980Ti in Oculus games: NVIDIA GTX 1080 Performance Review: Head to Head Against the 980 Ti - Page 3 of 3 - Road to VR
But [H] may be first to bench Raw Data, that I'm not disputing.
Click to expand...
My point was that PCPer did not do a comparison of GPUs but rather were comparing headset performance.

Thanks for that Road to VR link!!! I had not seen that. That is interesting in how he showed the data. He was using canned 2D benchmarks too, a VR movie, and Project Cars which I think was actual gameplay, not sure though, just worth noting here. That was the biggest hurdle on what we did here. It was not collecting the data, as obviously can be done easily, but rather how to display that data in a digestible and understandable fashion. Given that nobody has complained yet about our graphing, I guess we did OK.

That said, what do you guys think about keeping the graphs separate? Once you put two GPUs on a single graph, it gets messy quickly.
 
Kyle,

Would you say that VR is "mature enough" that a 1080 and decent supporting HW is enough to get value out it? I know that for the "novelty enthusiast" it has been a good place..but is it to the point of where larger audience adoption will say "yep, this is cool and worth the $$$ I spent"?

Reason I ask..is my PC is old, but I cannot find anything I play right now that justifies an "upgrade"...only thing I can think of is VR.
 
Trepidati0n said:
Kyle,

Would you say that VR is "mature enough" that a 1080 and decent supporting HW is enough to get value out it? I know that for the "novelty enthusiast" it has been a good place..but is it to the point of where larger audience adoption will say "yep, this is cool and worth the $$$ I spent"?

Reason I ask..is my PC is old, but I cannot find anything I play right now that justifies an "upgrade"...only thing I can think of is VR.
Click to expand...
Simple answer....yes. There are still some drawbacks to VR. You are likely not going to put in multi-hour gaming sessions just due to the headset. It gets heavy and hot after a while and it can be a bit cumbersome with the cables. I would highly suggest going and using a demo first or finding someone that is willing to share the experience with you.
 
On the subject of multi GPU, Unreal Engine uses deferred rendering. While not impossible to do AFR with deferred rendering, it is difficult and expensive. The question is if any of the new multiadapter implementations in DX12 will improve the situation.
 
You must log in or register to reply here.
Back
Top