[MERGED] DirectX 11/12, OpenGL, VulkanAPI Discussion

Hi all

 

We are wide awake thanks, Vulkan is being actively worked on. We indicated this in the 2020 road map.

 

It is a massive time consuming project, we hope to start internal testing towards the end of the year.

 

As soon as we can bring you news we will.

 

thanks

Then everything is said. Awesome thing!

Would be nice to have one Thread per AI if thats somehow possible :) That can scale nicely with Future cpu Gens.

Keep up the good work.

Gesendet von iPhone mit Tapatalk

Hi all

 

We are wide awake thanks, Vulkan is being actively worked on. We indicated this in the 2020 road map.

 

It is a massive time consuming project, we hope to start internal testing towards the end of the year.

 

As soon as we can bring you news we will.

 

thanks

Thanks BN. So when I´ll see implemented, hope by the end of year, I would consider buying any other modules, but not before. I wouldn´t even try channel maps, syria etc at 20 or less fps in VR, better no thanks.

"Transfering Shaders" does not mean Copy/Paste, they would need to be re-written.

 

A Little Recap for those that think DX11 -> Vulkan will not help with FPS.

Thank you.

Please ED, finally move in here. New modules, new maps, ... completely irrelevant.

Module teams working with the SDK and Maps teams working with TDK (Terrain develop Kit), has nothing to do with the Core team and DCS engine. Diferent competences and diferent tools and funtionality.

Thanks BN. So when I´ll see implemented, hope by the end of year, I would consider buying any other modules, but not before. I wouldn´t even try channel maps, syria etc at 20 or less fps in VR, better no thanks.

+1

Thanks BN. So when I´ll see implemented, hope by the end of year, I would consider buying any other modules, but not before. I wouldn´t even try channel maps, syria etc at 20 or less fps in VR, better no thanks.

+1

I always grin when I see these type of comments. Its like you guys are threatening ED that if they do not do as you say, they will go out of business or something. You guys go ahead and not buy anything and miss out on all the great DCS has to offer. You will not be missed.

I always grin when I see these type of comments. Its like you guys are threatening ED that if they do not do as you say, they will go out of business or something. You guys go ahead and not buy anything and miss out on all the great DCS has to offer. You will not be missed.

This quote sounds like a threat to you?

"Thanks BN. So when I´ll see implemented, hope by the end of year, I would consider buying any other modules, but not before. I wouldn´t even try channel maps, syria etc at 20 or less fps in VR, better no thanks."

Wierd.

I always grin when I see these type of comments. Its like you guys are threatening ED that if they do not do as you say, they will go out of business or something. You guys go ahead and not buy anything and miss out on all the great DCS has to offer. You will not be missed.

Paying £60+ per a single plane, it must definitely offer more than just that. Me, I'm not missing anything. DCS is not the last biscuit in packet.

Sort of a tangent, but I recently came back to DCS after being away from my DCS hardware for about nine months to a year. The improvements in VR all on the same hardware have been remarkable.

If you're on regularly with the openbeta sometimes it is hard to see the incremental performance stuff especially when bugs and issues that pop up. As far as the big picture, I'm encouraged by the last year's progress.

Sort of a tangent, but I recently came back to DCS after being away from my DCS hardware for about nine months to a year. The improvements in VR all on the same hardware have been remarkable.

 

If you're on regularly with the openbeta sometimes it is hard to see the incremental performance stuff especially when bugs and issues that pop up. As far as the big picture, I'm encouraged by the last year's progress.

+1

As a person with fairly low spec for VR (GTX 1060, 16GB) I have noticed each stable release over the past year has given a substantial performance improvement.

BZ ED!!

// MERGED

Hi all

 

We are wide awake thanks, Vulkan is being actively worked on. We indicated this in the 2020 road map.

 

It is a massive time consuming project, we hope to start internal testing towards the end of the year.

 

As soon as we can bring you news we will.

thanks

* this is very important to know, keep us informed please

* this is the reason, why I keep buying new modules and maps, even though I make jokes about the Channel performance; sorry for this, I had to. :)

* so I keep my fingers crossed!

Hi here, I have question about Vulkan:

What is it ?

Will it make the game run more smoothly ?

Will it change the way the game looks ? Like 1.5 to 2.0 was a massive improvment in the graphics and my old laptop could not support DCS 2.0 so I had to buy a new one.

I ask this because I think about buying a new, better PC, but if Vulkan requires better hardware, I will definitely wait until it's released before buying anything. :)

Hi here, I have question about Vulkan:

What is it ?

Will it make the game run more smoothly ?

Will it change the way the game looks ? Like 1.5 to 2.0 was a massive improvment in the graphics and my old laptop could not support DCS 2.0 so I had to buy a new one.

 

I ask this because I think about buying a new, better PC, but if Vulkan requires better hardware, I will definitely wait until it's released before buying anything. :)

Vulkan is newer technology. Usually newer means better.

Hi here, I have question about Vulkan:

What is it ?

Will it make the game run more smoothly ?

Will it change the way the game looks ? Like 1.5 to 2.0 was a massive improvment in the graphics and my old laptop could not support DCS 2.0 so I had to buy a new one.

 

I ask this because I think about buying a new, better PC, but if Vulkan requires better hardware, I will definitely wait until it's released before buying anything. :)

May be this post from SkateZilla will also give you some inside

https://forums.eagle.ru/showpost.php?p=4385787&postcount=221

Thanks a lot ! I understand better now.

I read many times on the hardware related section of the forum that it does not matter if your CPU has 8 cores or more, what matters is the speed of one core. This is explained by the way DirectX11 functions.

However, it will not be correct anymore with Vulkan. So with Vulkan, it will be better to have more cores on the CPU rather than having a higher frequency on less cores . I am correct ?

Thanks a lot ! I understand better now.

 

I read many times on the hardware related section of the forum that it does not matter if your CPU has 8 cores or more, what matters is the speed of one core. This is explained by the way DirectX11 functions.

However, it will not be correct anymore with Vulkan. So with Vulkan, it will be better to have more cores on the CPU rather than having a higher frequency on less cores . I am correct ?

Correct, although single core performance by itself will always be beneficial to gaming. Multi core among other things, help process data in the background and/or in parallel.

Thanks a lot ! I understand better now.

 

I read many times on the hardware related section of the forum that it does not matter if your CPU has 8 cores or more, what matters is the speed of one core. This is explained by the way DirectX11 functions.

However, it will not be correct anymore with Vulkan. So with Vulkan, it will be better to have more cores on the CPU rather than having a higher frequency on less cores . I am correct ?

In perfect world you are correct. In reality most (DX12/Vulkan) games still run better on CPU with less cores but higher singlecore speed

As i understand it , Vulkan and multithreading are two entirely different things .

The API will improve frametimes by unloading the CPU due to pre-cached shaders in the VRAM , making draw calls easier on the CPU .

The only hardware consideration when readying for Vulcan is more VRAM , to accommodate the pre-cache .

As i understand it , Vulkan and multithreading are two entirely different things .

The API will improve frametimes by unloading the CPU due to pre-cached shaders in the VRAM , making draw calls easier on the CPU .

The only hardware consideration when readying for Vulcan is more VRAM , to accommodate the pre-cache .

https://developer.nvidia.com/sites/default/files/akamai/gameworks/blog/munich/mschott_vulkan_multi_threading.pdf

Vulkan was designed from ground up for multi-core, multi-thread platforms and allows the generation of graphics commands in parallel to the commands execution while previous APIs don't. Another benefit is parallel buffer generation which makes better use of all CPU cores. But of course, Vulkan is great on a CPU-bound setting, but still most graphics demanding applications are still GPU-bound, so the performance improvement is limited, although necessary for what it does is to allow a better or full use of the GPU capacity.

Vulkan was designed from ground up for multi-core, multi-thread platforms and allows the generation of graphics commands in parallel to the commands execution while previous APIs don't. Another benefit is parallel buffer generation which makes better use of all CPU cores. But of course, Vulkan is great on a CPU-bound setting, but still most graphics demanding applications are still GPU-bound, so the performance improvement is limited, although necessary for what it does is to allow a better or full use of the GPU capacity.

Thanks for the link , very interesting .

I for one am very excited to see performance improvements with Vulkan. I don't play many other games on my computer besides DCS, but one game i have played is DOOM Eternal. I mention this because it uses Vulkan and although its not a simulator, it is graphically intensive with lots of stuff going on at any one time. On my system playing that in 2k at full maxed settings - nightmare settings - I get anywhere from 250-300fps - performance is outstanding. Again, its not a fair comparison, but its just a very positive example of an excellently performing title using Vulkan.