[David Kanter on Tech Report] - Nvidia VR preemption "possibly catastrophic".

[njdevilsfan87]

I'm certainly looking forward to VR. It should be a good time for some GPU upgrades then. So I'm not concerned about what current GPUs can or can't do with respect to VR.

 

[happy medium]

since they have likely taped out pascal, does anyone know if nv will have had enough time to fix the async compute issues with pascal or will they have to wait to do it with volta?

What? this is insane............AMD public relations at its best!

How many VR games can Nvidia gpu's not play? answer 0 because there is none.
and won't be any for quite a while.

 

[Zstream]

What? this is insane............AMD public relations at its best!

How many VR games can Nvidia gpu's not play? answer 0 because there is none.
and won't be any for quite a while.

I have a 270 -> 270x speeds. I would like my purchase to last a little bit

 

[antihelten]

What? this is insane............AMD public relations at its best!

How many VR games can Nvidia gpu's not play? answer 0 because there is none.
and won't be any for quite a while.

Actually there are plenty of VR games out there (although off the top of my head, there's only one commercial game that I can think off: Elite: Dangerous)

But Nvidia can of course play all of those without problems, since none of them are particularly demanding.

 

[Pariah]

The fact that we keep trying means we want it lol....
VR, 3D, etc.
Yes sure, some technologies will fail, but it's only getting more and more advanced, until eventually we get Caprica/Battlestar Galactica!

Movies have made VR look awesome to the general public. What people want is Ghost in the Shell. We are not remotely close to that.

for a forum of supposed tech enthusiasts, you all sure don't seem to have enthusiasm for anything new. All I see is complaining and talking every new tech down for no reason at all.

Our definitions of tech enthusiasts differ. My definition of a tech enthusiast is not someone who is a sucker for every fad, gimmick, and half baked alpha product that suffers from serious drawbacks because the technical building blocks necessary to make it a viable product have not been developed yet.

 

[gorobei]

What? this is insane............AMD public relations at its best!

How many VR games can Nvidia gpu's not play? answer 0 because there is none.
and won't be any for quite a while.

http://www.forbes.com/sites/paullamkin/2015/06/12/oculus-rift-consumer-edition-5-things-we-now-know/

4. There’s some heavy hitting gaming action coming

We’ve already seen plenty of cool VR gaming demos on the Oculus platform but the list of titles getting an E3 airing is pretty exciting.

EVE: Valkyrie from CCP Games, Chronos from Gunfire Games, Damaged Core by High Voltage, VR Sports: Challenge by Sanzaru, Esper from Coatsink, AirMech by Carbon, Lucky’s Tale from Playful and Edge of Nowhere from Insomniac Games will all be on show at the big LA gaming expo next week, and the company also announced a $10 million fund to accelerate indie developers.

5. The price and release date

Actually, what we discovered about the actual launch date and pricing of the Oculus Rift was pretty minimal. The company merely confirmed earlier reports that the headset will go on sale in the first quarter of 2016.

both valve and oculus have pushed back release from this year to Q1'16. so either 4 or 7 months from now, that is nothing in terms of time. its safe to assume there will be at least 1 to 3 games at release time.

according to TestedPodcast elite dangerous has ended Win10 compatibility patches with the older rift dk2 and are waiting for cv1 release, generally indicating that ED on rift cv1 will be soon enough that they are willing to risk pissing off the current ED vr players.

since nv apologists have indicated that any issues will be solved with the next gen of cards, that would require that nv increased priority on implementing AC and preemption during pascal planning in order for it to be in the next gen. according to the kanter interview the current deficiency is a result of prioritizing for the games at the time of maxwell and the issues with tsmc going to 20nm. they sacrificed compute for gaming power because there wasnt a process shrink to allow die space for compute.

since many here follow nv and their architecture far more closely than the general public, they should have a higher chance of knowing whether pascal is a balance of compute and gaming or another compromise.

if someone is looking to buy a next gen card for VR then this is a relevant question.(im looking so im asking).

 

[Erenhardt]

What? this is insane............AMD public relations at its best!

How many VR games can Nvidia gpu's not play? answer 0 because there is none.
and won't be any for quite a while.

There are a lot of VR games.
War thunder,
Assetto Corsa,
Alien Isolation,
EuroTruck Sim 2,
+ a lot of compatible games
+ a lot of early access games

 

[moonbogg]

None of this VR stuff matters to me at all. What VR? People talk about it like millions of gamers have VR headsets on their desks right now. I don't even know how to buy a VR headset, who makes them, which ones are good, or if they are even coming out in the near or far future. I have no idea. VR is as good as a myth to me.
Also, as far as my intuitions tell me, VR will be born, suffer a crappy life and die crying in a corner somewhere just like 3D did. 3D was painful to use with those terrible glasses, poor performance, 3D not working for many games and lagging etc etc. For VR to be good, you need to ditch the helmet and build a holodeck, because I simply do not see eye goggles, helmets or anything you must stick on your head of face going mainstream in any capacity.

 

[MongGrel]

Still haven't gotten all the way through this.

Is interesting though.

 

[BSim500]

Our definitions of tech enthusiasts differ. My definition of a tech enthusiast is not someone who is a sucker for every fad, gimmick, and half baked alpha product that suffers from serious drawbacks because the technical building blocks necessary to make it a viable product have not been developed yet.

^ This. There's a big difference between someone who's enthusiastic about the potential of new tech, vs an impatient hyper-consumerist who jumps on every new marketing driven fad and sneers down at anyone who wants it done right more than they want it done quickly. Right now VR isn't anywhere near "good enough" for the amount of hype ascribed to it. Almost no-one owns a high-end rig capable of driving multiple high res display, whilst many people find the lower res VR units are less pleasant than sitting in front of a decent monitor given how close they sit to the eyes. Still a lot of issues to work out : dislike of headsets, headset weight / comfort over several hours (beyond a 10-20min demo), comfort during hot weather, "VR sensor fusion" (eyes + inner ear positioning) only works on rotation axis not translation axis (beyond simple leaning a few inches) whilst sitting down, ongoing motion sickness even without any latency, spectacle wearer's discomfort, light bleed around the nose area, overnight grid persistence issues, calibration issues, can't see input device (keys on keyboard) so maybe limited to a controller which changes / limits gameplay to "fit" the device, headset based surround sound less accurate than a 4-speaker setup but speakers do not "rotate" with your head leading to confusion, etc. Some claim they can "push their way through" motion sickness, but how many consumer entertainment devices are going to sell on the back of "ignore the vomit, it'll go away after 18 months of training!" which sounds more like a 1950's Soviet nuclear fallout survival manual than a gaming enhancement... :biggrin:

I think it's going to take longer than VR firms wish simply because half the "problem" isn't just technology, it's that human beings are inherently different with different eyesight, tolerances and visual likes / dislikes. From what I've heard, even if you perfect rotation axis (pitch, roll, yaw), you cannot perfect the translation axis (x, y, z body movements) in the same way that maintain perfect nausea free simulation, because as soon as you start moving more than 1-2ft at a time (running / strafing / falling / FPS style jump pads) then your inner ears obviously cannot produce the "motion" that matches what you see (and which your brain now has a greater expectation of feeling vs a 2D monitor given the accurate rotation simulation), then a number of people get motion sick again from the "split" that occurs whilst you physically remain stationary when only the rotation axis movements will have an accompanying inner ear body movement, but not translation movements (beyond leaning a few inches). Even if you manufactured some kind of 2D walking treadmill, A: It still couldn't cope with accurate portrayal of sudden vertical movement (falls, jump pads, etc), and B: That simply doesn't fit in with the psychology of how most gamers play (flop out on the couch / bed / chair to relax).

Likewise, there's more to visual discomfort than just motion sickness. In real life your eyes work to both focus and converge on a point in space (Accommodation-Convergence Reflex). Since the focus/converge distance is the same, your brain has since birth learned to "couple" the two response together (Vergence-Accommodation Coupling). Headsets completely break that natural reflex as your eyes will be focal locked to only one distance (eyeball to VR screen distance) whilst your brain has an opposing instinct of variable distance convergence when "tricked" with the 3D effect. This "Vergence-Accommodation Conflict" isn't motion sickness, it's an additional effect on top of that which a lot of people find plain unpleasant for more than a few minutes / seconds, causes eye-strain, headaches, etc, even with silky smooth jitter free 120fps. So there's far more to "VR comfort" than just "stuff up the fps to 90-120, eliminate the jitter and all will be well". The very nature of how headsets try and split the natural focal/convergence paired reflex coupling, plus general eyestrain for many of constant forced focus of mere inches away is actually extremely unnatural as to how human vision works in reality.

By the time "full ergonomic total immersion VR" is "properly" ready (which may not even involve headsets in the long run), yes every current GFX card and every current VR kit will long be as totally obsolete as the 1st gen 1x speed Mitsumi clamshell CD-ROM drives were when 8-16x speed DVD-burners came out. Everything in between now and that is really just varying stages of Alpha/Beta testing for the 0.07% of the PC market who even care about enhanced realism enough to buy a single 2D 4K screen today.

 

[Mondozei]

^ This. There's a big difference between someone who's enthusiastic about the potential of new tech, vs an impatient hyper-consumerist who jumps on every new marketing driven fad and sneers down at anyone who wants it done right more than they want it done quickly. Right now VR isn't anywhere near "good enough" for the amount of hype ascribed to it. Almost no-one owns a high-end rig capable of driving multiple high res display, whilst many people find the lower res VR units are less pleasant than sitting in front of a decent monitor given how close they sit to the eyes. Still a lot of issues to work out : dislike of headsets, headset weight / comfort over several hours (beyond a 10-20min demo), comfort during hot weather, "VR sensor fusion" (eyes + inner ear positioning) only works on rotation axis not translation axis (beyond simple leaning a few inches) whilst sitting down, ongoing motion sickness even without any latency, spectacle wearer's discomfort, light bleed around the nose area, overnight grid persistence issues, calibration issues, can't see input device (keys on keyboard) so maybe limited to a controller which changes / limits gameplay to "fit" the device, headset based surround sound less accurate than a 4-speaker setup but speakers do not "rotate" with your head leading to confusion, etc. Some claim they can "push their way through" motion sickness, but how many consumer entertainment devices are going to sell on the back of "ignore the vomit, it'll go away after 18 months of training!" which sounds more like a 1950's Soviet nuclear fallout survival manual than a gaming enhancement... :biggrin:

I think it's going to take longer than VR firms wish simply because half the "problem" isn't just technology, it's that human beings are inherently different with different eyesight, tolerances and visual likes / dislikes. From what I've heard, even if you perfect rotation axis (pitch, roll, yaw), you cannot perfect the translation axis (x, y, z body movements) in the same way that maintain perfect nausea free simulation, because as soon as you start moving more than 1-2ft at a time (running / strafing / falling / FPS style jump pads) then your inner ears obviously cannot produce the "motion" that matches what you see (and which your brain now has a greater expectation of feeling vs a 2D monitor given the accurate rotation simulation), then a number of people get motion sick again from the "split" that occurs whilst you physically remain stationary when only the rotation axis movements will have an accompanying inner ear body movement, but not translation movements (beyond leaning a few inches). Even if you manufactured some kind of 2D walking treadmill, A: It still couldn't cope with accurate portrayal of sudden vertical movement (falls, jump pads, etc), and B: That simply doesn't fit in with the psychology of how most gamers play (flop out on the couch / bed / chair to relax).

Likewise, there's more to visual discomfort than just motion sickness. In real life your eyes work to both focus and converge on a point in space (Accommodation-Convergence Reflex). Since the focus/converge distance is the same, your brain has since birth learned to "couple" the two response together (Vergence-Accommodation Coupling). Headsets completely break that natural reflex as your eyes will be focal locked to only one distance (eyeball to VR screen distance) whilst your brain has an opposing instinct of variable distance convergence when "tricked" with the 3D effect. This "Vergence-Accommodation Conflict" isn't motion sickness, it's an additional effect on top of that which a lot of people find plain unpleasant for more than a few minutes / seconds, causes eye-strain, headaches, etc, even with silky smooth jitter free 120fps. So there's far more to "VR comfort" than just "stuff up the fps to 90-120, eliminate the jitter and all will be well". The very nature of how headsets try and split the natural focal/convergence paired reflex coupling, plus general eyestrain for many of constant forced focus of mere inches away is actually extremely unnatural as to how human vision works in reality.

By the time "full ergonomic total immersion VR" is "properly" ready (which may not even involve headsets in the long run), yes every current GFX card and every current VR kit will long be as totally obsolete as the 1st gen 1x speed Mitsumi clamshell CD-ROM drives were when 8-16x speed DVD-burners came out. Everything in between now and that is really just varying stages of Alpha/Beta testing for the 0.07% of the PC market who even care about enhanced realism enough to buy a single 2D 4K screen today.

This is such a great post, I wish I could give it gold á la Reddit.

 

[DeathReborn]

I have never been a big fan of VR due to the nausea it gives me, despite that I keep trying new models and get the same results. I do thin k some people see the term VR and think Holodeck and what we have coming soon/now is nothing like that.

I think this VR push will fail die to a lack of market penetration no matter how well it performs, people en mass will not buy it so prices will remain high and slowly fade away.

My upgrade cycle is 18 months to 2 years (will probably be 3 years by decades end) so not having a future proof architecture (mostly using 290's now...) doesn't bother me.

 

[Mondozei]

I have never been a big fan of VR due to the nausea it gives me, despite that I keep trying new models and get the same results. I do thin k some people see the term VR and think Holodeck and what we have coming soon/now is nothing like that.

I think this VR push will fail die to a lack of market penetration no matter how well it performs, people en mass will not buy it so prices will remain high and slowly fade away.

My upgrade cycle is 18 months to 2 years (will probably be 3 years by decades end) so not having a future proof architecture (mostly using 290's now...) doesn't bother me.

I'm ultimately a believer in VR, but I am not as fanatically idealistic about VR as I was, say, 6 months ago.

Truly immersive VR will make people nauseous. There was a great comment in one of the recent GDC talks on VR. A dev relayed a story where he'd get complaints about people getting nauseous from playing a jet game.

And he asked, rhetorically: well, have you ever flown a real jet before?
Point is, once VR gets really good, the thing that gets us sick IRL will likely get us sick in VR as well.

For this reason alone, fast-paced shooters will likely be much more difficult to do in VR and why I think monitors will continue to have a long shelf life rather than the instant death many predict. Over time I expect VR to fully take over but the transition will probably take a lot longer than many VR enthusiasts like myself predicted just a few months ago.

 

[3DVagabond]

I'm ultimately a believer in VR, but I am not as fanatically idealistic about VR as I was, say, 6 months ago.

Truly immersive VR will make people nauseous. There was a great comment in one of the recent GDC talks on VR. A dev relayed a story where he'd get complaints about people getting nauseous from playing a jet game.

And he asked, rhetorically: well, have you ever flown a real jet before?
Point is, once VR gets really good, the thing that gets us sick IRL will likely get us sick in VR as well.

For this reason alone, fast-paced shooters will likely be much more difficult to do in VR and why I think monitors will continue to have a long shelf life rather than the instant death many predict. Over time I expect VR to fully take over but the transition will probably take a lot longer than many VR enthusiasts like myself predicted just a few months ago.

Typically what causes "motion sickness" is conflicting information between our eyes and inner ear. Seems like there isn't really any way to change that with VR. Some people will be bothered and some won't. Some will get used to it and get over the nausea, some won't. In the end this could be a show stopper for VR if enough people suffer from it.

 

[Sabrewings]

I have a few dozen hours in Elite in VR, and I can tell those in this thread who haven't used it. The experience is difficult to convey, but the feeling of presence is irreplaceable and addictive. I have also had a few other experiences with the Oculus DK2 and it is absolutely clear to me that it will be a hit. I see the same uninformed and useless comparisons to 3D, but 3D TVs/monitors have nothing on VR for the experience of presence.

Everyone I know who has tried it is now a believer. You just don't forget being so seamlessly moved to another world. Yes, inner ear issues could be an issue for some, but games like flight sim and racing thus far translate very well.

Regarding the whole room experiences, find a bad review for the HTC Vice, I dare you.

 

[Mondozei]

Typically what causes "motion sickness" is conflicting information between our eyes and inner ear. Seems like there isn't really any way to change that with VR. Some people will be bothered and some won't. Some will get used to it and get over the nausea, some won't. In the end this could be a show stopper for VR if enough people suffer from it.

Right, but we got to distinguish between VR nausea that is induced because of bad implementation and VR nausea that is there simply because what you are simulating is going to get (some) people sick, VR or no VR.

My point was concentrated on the latter part. That no matter how good we get at VR, some things will always get some people sick.

Think of how tolerant people are of rollercoasters, for example. These things are very individual. The exact same rules will apply in VR. That makes it a harder medium to code and develop for than for desktop gaming, even if you do everything right from a technical standpoint.

For this reason I think VR will take off more slowly than many of us thought in the first place. I certainly remain a strong VR believer in the long run but I probably underestimate just how hard it can be to develop for VR because of these constraints. Normally on desktop PC gaming, if your game is full of bad code it will perform badly. Likewise, if your code is great and optimise you don't have to worry.

With VR you have this extra layer where your code could be awesome but simply the nature of the experience can make some people quite sick. That is something new entirely and an added challenge.

Another point is that a lot of devs for VR become accustomed and develop a sense of VR acclimatisation. So it's a lot harder to get a more accurate sense of how a first-time VR user would experience something. You can't just assume because you yourself and your team don't get sick, nobody else will, especially if you're in VR every day of the work week.

 

[railven]

I have a few dozen hours in Elite in VR, and I can tell those in this thread who haven't used it. The experience is difficult to convey, but the feeling of presence is irreplaceable and addictive. I have also had a few other experiences with the Oculus DK2 and it is absolutely clear to me that it will be a hit. I see the same uninformed and useless comparisons to 3D, but 3D TVs/monitors have nothing on VR for the experience of presence.

Everyone I know who has tried it is now a believer. You just don't forget being so seamlessly moved to another world. Yes, inner ear issues could be an issue for some, but games like flight sim and racing thus far translate very well.

Regarding the whole room experiences, find a bad review for the HTC Vice, I dare you.

I don't doubt that people that use it will be impressed or like it. But, just list out the cost of entrance for your experience.

Then factor in Joe and Jane gamer aren't going to have that cash lying around. Then on the console side, it has to get deep support for it to even survive (techs die on consoles as niche/fads so quickly. THQ died to their stupid Paint it game thingie bob who's name I can't even remember!)

 

[3DVagabond]

I thought what made their APU's for consoles so good was both Intel and Nvidia said no.

One of the requirements was it be X86. nVidia isn't even in that picture. Intel might have been for too low profit though. AMD had more to gain though, so it explains why it would be worthwhile for them. Plus, I don't think they would want to turn down the business. It's not like they do the volume that Intel does.

 

[positivedoppler]

I have never been a big fan of VR due to the nausea it gives me, despite that I keep trying new models and get the same results. I do thin k some people see the term VR and think Holodeck and what we have coming soon/now is nothing like that.

I think this VR push will fail die to a lack of market penetration no matter how well it performs, people en mass will not buy it so prices will remain high and slowly fade away.

My upgrade cycle is 18 months to 2 years (will probably be 3 years by decades end) so not having a future proof architecture (mostly using 290's now...) doesn't bother me.

I have the opposite problem. I get motion sickness playing almost all 3d game, but I'm fine with VR. I've waited for thus for sooooo long

 

[Dribble]

I have never been a big fan of VR due to the nausea it gives me, despite that I keep trying new models and get the same results. I do thin k some people see the term VR and think Holodeck and what we have coming soon/now is nothing like that.

The closest thing to the holodeck isn't a headset, it's a cave (i.e. a special room which all the walls are back projected stereo screens). Then you just wear 3d glasses and the virtual environment seems to envelop you. These have been around since the 90's too and are more popular then headsets because they work better - e.g. some engineering companies really use the things to help design cars. Unfortunately caves are very expensive.

 

[Gikaseixas]

Personally I dont see a future for VR outside the niche segment.

For AMD its always the next thing around the corner. And we all know how reality plays that game.

If OP didn't have Nvidia painted in a negative way many here would not claiming "VR doom" before it's arrival, plus VR isn't a AMD tech, it's just happens that AMD's implementation "might" be the better one so tone down your hatred dude

 

[sontin]

If OP didn't have Nvidia painted in a negative way many here would not claiming "VR doom" before it's arrival, plus VR isn't a AMD tech, it's just happens that AMD's implementation "might" be the better one so tone down your hatred dude

There is no comparision between nVidia and AMD so Kanter can say whatever he want. Nobody can verify it.

On the other hand HTC is promoting nVidia and the Open Source Virtual Reality has implemented nVidia's GameworksVR libary.

 

[DeathReborn]

I'm ultimately a believer in VR, but I am not as fanatically idealistic about VR as I was, say, 6 months ago.

Truly immersive VR will make people nauseous. There was a great comment in one of the recent GDC talks on VR. A dev relayed a story where he'd get complaints about people getting nauseous from playing a jet game.

And he asked, rhetorically: well, have you ever flown a real jet before?
Point is, once VR gets really good, the thing that gets us sick IRL will likely get us sick in VR as well.

For this reason alone, fast-paced shooters will likely be much more difficult to do in VR and why I think monitors will continue to have a long shelf life rather than the instant death many predict. Over time I expect VR to fully take over but the transition will probably take a lot longer than many VR enthusiasts like myself predicted just a few months ago.

I am actually qualified to fly most aircraft up to small jets so flying for me is not an issue, it's the loss of peripheral vision and the disconnect in the audio-visual synchronisation that gets to me.

 

[Glo.]

About VR.

https://twitter.com/missquickstep/status/637478111957815296

Here you go.

 

[sontin]

Zlatan isn't saying anything that anyone following VR closely doesn't already know. If people really understood how far ahead AMD is in VR they'd be stunned. There is even bigger news to come from AMD and Crytek. Nvidia doesn't have the hardware or software, it's like a 286 on DOS vs an i7 on Windows 10.

Ha, you're absolutely right: https://www.youtube.com/watch?v=zFCxjXe3f1M

Oh and here is nVidia with timewarps: https://www.youtube.com/watch?v=uwGDg_SegDg

12ms latency.