Some nice VR Improvement coming!

[Shack]

Great news !

[pimp]

Wow this is great news. No need for vulkan now.:lol:

[Svsmokey]

Wow this is great news. No need for vulkan now.:lol:

 

Stop that ! :)

[Willie Nelson]

Can anyone explain in layperson’s terms what the difference between these terrain fixes and a potential Vulcan api are?

 

Not to be too sceptical of Wags but I’m confused. If it’s been a terrain issue in large part then should we not have expected excellent performance in Bluewater territory?

[Holyroller1]

BOOM goes the dynamite!!!

[The Falcon]

Can anyone explain in layperson’s terms what the difference between these terrain fixes and a potential Vulcan api are?

Not to be too sceptical of Wags but I’m confused. If it’s been a terrain issue in large part then should we not have expected excellent performance in Bluewater territory?

do you notice any difference between flying over the earth instead of flying over the water in vr?

[Gman109]

What great timing, thanks Wags/ED. I've been constantly upgrading my VR units, first the Vives/Rifts, then to the Odyssey, then the Vive Pro/Odyssey+, now the Pimax5k...Cannot WAIT for the new upcoming VR units now with this news.

 

 

Hopefully this new performance increase courtesy of Wags/ED will have a good impact on the FPS/capability of new units like the HP one etc.

[Svsmokey]

I'm wondering what the effects will be cpu vs gpu . Will this take my 1070 to 2080 passmark or my 6500 to stock-clock 9900k performance ? I'll take either of the these :) But as the answer will impact my upgrading plans , i would like to know...

Edited April 5, 2019 by Svsmokey

[Tinkickef]

Can anyone explain in layperson’s terms what the difference between these terrain fixes and a potential Vulcan api are?

 

Not to be too sceptical of Wags but I’m confused. If it’s been a terrain issue in large part then should we not have expected excellent performance in Bluewater territory?

 

It would seem that since the terrain engine is affecting VR performance specifically, it is introducing latency in GPU rendering of the binocular view of the two screens. This has badly affected performance.

 

Vulkan allows the game to utilise all cores of a CPU instead of just one (and a bit for audio on another core). If you have a quad core CPU, this will effectively use the currently redundant cores giving a huge increase in computational power, reducing lag and allowing all the bells and whistles to be turned up right to the full capability of the GPU.

[nibbio]

Can anyone explain in layperson’s terms what the difference between these terrain fixes and a potential Vulcan api are?

 

Not to be too sceptical of Wags but I’m confused. If it’s been a terrain issue in large part then should we not have expected excellent performance in Bluewater territory?

 

 

My guess is that terrain is still rendered underwater, even if not visible to the player.

[Willie Nelson]

It would seem that since the terrain engine is affecting VR performance specifically, it is introducing latency in GPU rendering of the binocular view of the two screens. This has badly affected performance.

 

Vulkan allows the game to utilise all cores of a CPU instead of just one (and a bit for audio on another core). If you have a quad core CPU, this will effectively use the currently redundant cores giving a huge increase in computational power, reducing lag and allowing all the bells and whistles to be turned up right to the full capability of the GPU.

 

Ok but this is a terrain fix only at this stage AFAIK. Can we imply that Vulcan is likely to come later?

 

I’m currently only pushing a CV1 off a 2080ti there is no problems worth worrying about but there is more likely to be on one of the newer HMD’s even with this shader mod, I guess?

[ElementLT]

with the engine being worked out do you guys think this will also all us to see targets more clearly. Maybe we can do away with Dot Only labels for some. OR does this has more to do with the PD?

Edited April 5, 2019 by ElementLT

[eaglecash867]

My guess is that terrain is still rendered underwater, even if not visible to the player.

 

I haven't really been keeping up with all of the developments in DCS along the way, but at one time, the water was the base layer of the terrain engine. Not sure if that's still true.

 

As far as Vulkan goes, we've all seen that subject discussed on these forums for years and it hasn't been implemented yet. My guess on that would be that in this particular application, its not being found to yield a big enough performance increase to warrant further development and implementation, with all of the risks associated with a major programming change.

[Fri13]

Can anyone explain in layperson’s terms what the difference between these terrain fixes and a potential Vulcan api are?

 

Not to be too sceptical of Wags but I’m confused. If it’s been a terrain issue in large part then should we not have expected excellent performance in Bluewater territory?

 

The vulcan is best explained with the AMD own representation:

 

If we can see a 50% improvement at the best, and then 15-20% at the normal situation, that even would be great.

 

Like think that there are people who fall below 45 FPS (even less who reach the 90 FPS) and gets like a 30-32 FPS. For them the 50% increment means they will reach that 45 FPS, a totally smooth VR experience without some odd warping at some positions.

 

Now if the vulkan is there to assist that even more, so we can use more of our CPU, and as DCS is very CPU heavy, we could expect to see more likeness less stuttering in missions with a lot of AI units.

 

And if we can expect to see other VR improvements in the hardware, like now the VR doesn't support SLI, a dual 3D cards rendering. But if we could get a change to see a SLI support where each eye is drawn by a own dedicated 3D card? That could be crazy improvement. No need to buy then expensive 1000€ 3D cards but you could really well get along with a two 200-300€ cards as you would get around 70-130 FPS at full HD, so easily meet the 45 FPS for both eyes.

[Fri13]

My guess is that terrain is still rendered underwater, even if not visible to the player.

 

That I think would be odd, as even at the end of 90's the new graphical features (don't now recall the name of tech) was that only a visible mesh and textures were rendered, and this allowed a 16 MB 3D cards (like Riva TNT) to perform excellent manner when workload was heavily cut as unseen parts were not drawn at all behind other models.

 

So why would there be even a water modeled under a terrain, that is invisible to begin with?

[nibbio]

So why would there be even a water modeled under a terrain, that is invisible to begin with?

 

 

Beats me, but I took a couple of plunges off the edge of the carrier with the Tomcat, and the sea bottom was visible all around for miles...

Very clear "water" btw

[speed-of-heat]

also water "seems" to be modelled under the land as well...or at least it is rendered first, I dont know if it is updated after the land overlays it...

[Tinkickef]

Ok but this is a terrain fix only at this stage AFAIK. Can we imply that Vulcan is likely to come later?

 

I’m currently only pushing a CV1 off a 2080ti there is no problems worth worrying about but there is more likely to be on one of the newer HMD’s even with this shader mod, I guess?

 

It is a terrain engine fix. The terrain engine can be thought of as similar to the mission editor we plebs use. It allows the user to input terrain features, it is not the terrain itself, it is the underlying program that oversees stuff like scale, speed over ground and pretty much everything you see.

[Willie Nelson]

It is a terrain engine fix. The terrain engine can be thought of as similar to the mission editor we plebs use. It allows the user to input terrain features, it is not the terrain itself, it is the underlying program that oversees stuff like scale, speed over ground and pretty much everything you see.

 

Ok, I see. I guess we’ll find out more soon in any case. The proof will be in the pudding and perhaps Vulcan will not be relevant for a while. Appreciate your help.

[etherbattx]

Some nice VR Improvement coming!

 

Vulkan allows the game to utilise all cores of a CPU instead of just one (and a bit for audio on another core). If you have a quad core CPU, this will effectively use the currently redundant cores

 

if everything can be split between the cpu cores, are looking at a 400% increase with old processors and 800+% for some of the newer cpu’s?

[DerekSpeare]

This is good news! Thanks Wags!

[speed-of-heat]

if everything can be split between the cpu cores, are looking at a 400% increase with old processors and 800+% for some of the newer cpu’s?

 

That seems an unlikely expectation:

 

- ED would have to be 100% efficient in their re-architecture of their code base, at a theoretical level

 

- the operating system is still doing "things"... so that will take cycles away from Ed's code

 

- cpu thread movement is rarely 100% efficient

 

- as the cpu is being loaded more effectively, it is more likely to produce thermal throttling on the cpu, and similarly on the more effectively loaded GPU

 

 

as a result I would be ecstatic with 100% real world improvement, but I suspect 80% is more likely on a four physical core cpu, with reducing returns for 6 an 8 cores, my guess is max 120% and 130% over what we have today... (not including any code optimisations that Wags was talking about in this thread) … my 0.02$

[boedha68]

That seems an unlikely expectation:

 

- ED would have to be 100% efficient in their re-architecture of their code base, at a theoretical level

 

- the operating system is still doing "things"... so that will take cycles away from Ed's code

 

- cpu thread movement is rarely 100% efficient

 

- as the cpu is being loaded more effectively, it is more likely to produce thermal throttling on the cpu, and similarly on the more effectively loaded GPU

 

 

as a result I would be ecstatic with 100% real world improvement, but I suspect 80% is more likely on a four physical core cpu, with reducing returns for 6 an 8 cores, my guess is max 120% and 130% over what we have today... (not including any code optimisations that Wags was talking about in this thread) … my 0.02$[/quote

Well i think that Matt did not mention a false prediction. :D

[etherbattx]

one of the reasons vulkan is more performant is no more locking and synchronizing in dx11.

 

but, we sometimes forget that locking and synchronizing still needs to happen... the difference is that the game app gets to do al the work instead of dx11.

 

i would be surprised if we get much more than a 50% bump in FPS over dx11.

[Willie Nelson]

one of the reasons vulkan is more performant is no more locking and synchronizing in dx11.

 

but, we sometimes forget that locking and synchronizing still needs to happen... the difference is that the game app gets to do al the work instead of dx11.

 

i would be surprised if we get much more than a 50% bump in FPS over dx11.

 

Sounds like a good start, I’m pretty happy. I’m guessing we’ll need it for the Reverb, Cosmos, Pimax and whatever else may come out this year.