in development Vulkan API Discussion

I would also like that all developers at ED would only work on Multicore & Vulcan, but thats not working because this will not pay their salarys. But modules do. So they unfortenatly have to split it up.

Anyways providing some basic updates once per quarter should be possible. And I also think there is more information available as "we are working on it." What about some KPIs like "how many lines of code have been reworked / renewed / added" or "how many bugs have been encountered already"
Or as a really cool topic for a newsletter: An interview with one of the developers working on multicore.

Great idea. A Q&A with the dcs core developers... ( dynamic campaign included) 

5 hours ago, LucShep said:

I think everybody can understand and agree that things here work differently, but sometimes it gets to a point where things are ultimately questionable and, quite frankly, sometimes borderline distasteful.

Development on Vulkan engine for performance improvements was (AFAIK) announced nearly five years ago (DCS World Newsletter in 29 December 2017).

Five years.
During this time period, performance hasn't improved, it actually become even worse.

Noone asked for 2.7 higher performance impact, when 2.5 was already severe enough on hardware.
For those unwilling to accept it, there used to be 1.58 and 2.56 (legacy) versions available. They're gone. But even if available... great, if using that, then I'm blocked from the updates for the modules that I've paid for, and noone to play online...
And I don't care about the few people defending ED with "I play fine on 2.7 with six plus year old hardware", because that's just having far lower standards and accepting mediocre performance.
Paying customers, who have thrown quite significant ammounts into this flight/combat simulation, are forced to throw even bigger ammounts into hardware just to keep up with increasing resource demands, to get updates on purchased modules and play with others online. It's just not right. 

Meanwhile, five years on, little to absolutely no news on the Vulkan subject, while we keep having announcements upon announcements on new modules... which may (or not) have even bigger impact on performance than those that we currently have (did anyone even stop to think about this possibility?).

I think my point here is, the lack of perspective and the desproportion on how different areas seem to be focused on (IMO). 
Talking to people online, reading this and other forums, what I get is that there seems to be far more interest in "a better DCS" rather than in "more content in DCS".

Performance improvements on DCS should be the #1 priority, and sharing info on Vulkan development - no matter how small those news would be - would go a long way, and maybe should be considered as important as the constant "WIP" screens on new fancy announced modules (taking years to see a release) that are pumped weekly. 

 

You do of course realize there is another thread blowing up because the developers are announcing future projects “too soon”!????

Honestly its just complaining to complain…..

7 hours ago, LucShep said:

I think everybody can understand and agree that things here work differently, but sometimes it gets to a point where things are ultimately questionable and, quite frankly, sometimes borderline distasteful.

Development on Vulkan engine for performance improvements was (AFAIK) announced nearly five years ago (DCS World Newsletter in 29 December 2017).
....
Meanwhile, five years on, little to absolutely no news on the Vulkan subject, while we keep having announcements upon announcements on new modules... which may (or not) have even bigger impact on performance than those that we currently have (did anyone even stop to think about this possibility?).

I think my point here is, the lack of perspective and the desproportion on how different areas seem to be focused on (IMO). 
Talking to people online, reading this and other forums, what I get is that there seems to be far more interest in "a better DCS" rather than in "more content in DCS".

Performance improvements on DCS should be the #1 priority, and sharing info on Vulkan development - no matter how small those news would be - would go a long way, and maybe should be considered as important as the constant "WIP" screens on new fancy announced modules (taking years to see a release) that are pumped weekly. 

 

Delete or stop all "Fancy announced modules" never change nothing about improve a simulator engine develop, when the 3rd parties or the ED module studios has none about improve a graphic engine or dedicated personal with help them.

You need specialized personal, Engine programmer trained on game engine depeloping with understand the problems to implement Vulkan enginer into DCS engine, no a Sound, VFX, AI or database enginer or a 3D modeler or a  module builder. Use other resourse to improve them without training, has a waste of resourses, time and personal.

2 hours ago, Mr. Big.Biggs said:

You do of course realize there is another thread blowing up because the developers are announcing future projects “too soon”!????

Honestly its just complaining to complain…..

3rd party developers company studios, outside ED, and none to do with the ED core engine team?.

Luc>
the point here is that we know things are underway and comments like yours are not going to make any difference to the speed of delivery.

The only real effect is to waste yours and others time reading and responding to this and similar threads, whilst also potentially raising the blood pressure of some.

Wouldn’t your time be more enjoyably spent actually flying?

Hi all,

It's my personal opinion but getting disappointing that:

The first announcement of Vulcan happened at Nov 28, 2017.

As customer I just experience that there is no god to push enough performance out with VR from the game (without 3rd party tools). Before someone just asking I use: Ryzen 9 5900XT with ASUS 3090 OC and 64GB RAM, and yes I have spent enough time to find the best options.

If sometimes, still means that lots of already expected 'magic' still missing, like shadows generated by anything except the sun, moon and the 'tricky magic' shadows on carriers or

just a few annoying bugs:

- falling red colored snow,

- the fog still look like in the games around 2000 especially in rain,

- you hear the rain but not see it when using rain preset.

- after moved some LUAs into DLLs, the loading in menu became horribly slow,

- Mig29 cockpit 'unusable' at night since 2 years (no reaction for requests for fix especially if it has affect for ED modules)

- no fog visible in rain but rain moving into opposite direction than the wind,

- more then 90% of tracks not usable (We don't even dream that get a player where possible to forwards or backwards the time - in 2022 it is unthinkable),

- snow falling in hangar too,

and so on, many other things on 3rd party side as well.

At the same time 1-2 core of CPU just 'melting down' to serve everything, especially in multiplayer or missions, so sometimes some info or update would be enough to not sell the whole thousands of dollars worth equipment, and start to believe that something happening beyond the promises.

I quote again: "The first announcement of Vulcan happened at Nov 28, 2017."

34 minutes ago, 59th_Krieger said:

Hi all,

It's my personal opinion but getting disappointing that:

The first announcement of Vulcan happened at Nov 28, 2017.

As customer I just experience that there is no god to push enough performance out with VR from the game (without 3rd party tools). Before someone just asking I use: Ryzen 9 5900XT with ASUS 3090 OC and 64GB RAM, and yes I have spent enough time to find the best options.

If sometimes, still means that lots of already expected 'magic' still missing, like shadows generated by anything except the sun, moon and the 'tricky magic' shadows on carriers or

just a few annoying bugs:

- falling red colored snow,

- the fog still look like in the games around 2000 especially in rain,

- you hear the rain but not see it when using rain preset.

- after moved some LUAs into DLLs, the loading in menu became horribly slow,

- Mig29 cockpit 'unusable' at night since 2 years (no reaction for requests for fix especially if it has affect for ED modules)

- no fog visible in rain but rain moving into opposite direction than the wind,

- more then 90% of tracks not usable (We don't even dream that get a player where possible to forwards or backwards the time - in 2022 it is unthinkable),

- snow falling in hangar too,

and so on, many other things on 3rd party side as well.

 

At the same time 1-2 core of CPU just 'melting down' to serve everything, especially in multiplayer or missions, so sometimes some info or update would be enough to not sell the whole thousands of dollars worth equipment, and start to believe that something happening beyond the promises.

I quote again: "The first announcement of Vulcan happened at Nov 28, 2017."

 

It's been a journey I'll give you that, a graphics engine is not a simple task, add in the fact that the DCS Core is also being re-written to use Multi-Threading,
There's a lot of things that need time to make sure they are done correctly and that both Multi-Thread and Vulkan Transitions happen smoothly and they both work together.
 

3 minutes ago, NineLine said:

Not meaning to be blunt, but it will take the amount of time it needs to take. Again, I am not trying to be rude, but we have no reason to slow down or sit on these improvements. I dont know to what end it would help us by deliberately stalling any improvements of any kind. The last 3-4 years up till even today have been history making events that have turned the lives of many people upside down, and this all continues today and into tomorrow and the foreseeable future. On top of all that look at what we are trying to do, taking a very complex game and updating it on the fly with new tech that the game was never written to support, all the while adding new content and value for users to enoy AND trying to keep up with hardware demands.

So long story short, we are doing our best. These changes are now being tested internally but even this will take time.

I HIGHLY disagree with your second line, unless you have not bought one single improvement for your computer in 5 years, then maybe things have become worse.

Thanks.

We appreciate you came forward to us, although at the moment we look like an angry mob.

As a SW developer I usually renew my personal devices at period of at least 4 years. This means I can pretty much afford to buy PC which offers the decent performance. Sizable part of customer base can afford this over prolonged periods of time, and some do it actually every year.

However, many of your customers can not afford frequent hardware updates at the frequency the professionals do, far less than developers in gaming company. This means that a number of these users are affected heavily when ever you decide to leverage additional features ( generally desired) vs Moor's law (results in higher costs).

I would say I am quite low average at owning 1 map, FC3, Mig-21, F-16 and Ka-50. Notice how little the software costs compared to the needed hardware to run this thing. I would like to point out that investing in performance is not a wasted effort, it lets your user base be more elastic in terms of hardware ranges. Better performance with aged hardware means, that these people will remain with DCS the longer time. At the same time they may spend more money on your products than on expensive hardware updates. Please consider using this arguments when encountering Sales&Marketing people.

In short: the more we spend on NVIDIA/Intel/AMD, the less we spend on your products.

On 9/17/2022 at 5:57 PM, NineLine said:

Not meaning to be blunt, but it will take the amount of time it needs to take. Again, I am not trying to be rude, but we have no reason to slow down or sit on these improvements. I dont know to what end it would help us by deliberately stalling any improvements of any kind. The last 3-4 years up till even today have been history making events that have turned the lives of many people upside down, and this all continues today and into tomorrow and the foreseeable future. On top of all that look at what we are trying to do, taking a very complex game and updating it on the fly with new tech that the game was never written to support, all the while adding new content and value for users to enoy AND trying to keep up with hardware demands.

So long story short, we are doing our best. These changes are now being tested internally but even this will take time.

I HIGHLY disagree with your second line, unless you have not bought one single improvement for your computer in 5 years, then maybe things have become worse.

Thanks.

I did upgrade my computer within the last two years, not just but also on purpose for DCS (specs are in my sig) with only an archive HDD and SSD carried over.
By the time I finally had DCS 2.56 running sweet as a nut, 2.7 was out and it has been hell ever since.
Like others here, and as before, I had to decrease settings to a point where the game became pretty awfull compared to previous 2.56 version, doesn't look as good (because of reduced settings) or feel as good (because of unstable framerate).
So it has been a regression instead of progression. It, again, needs an even more potent GPU (€€€€). "I can't, I've invested enough already", I have to constantly remind myself. 

My point was more that any communication and feedback/news about performance improvements are rather welcome, it would be better than the current aproach of not saying anything. That can easily be interpreted as nothing being done, or even that something is going very wrong. Or felt as insecure as the unfortunate "yes, yes, two weeks be sure" that some master developer once said (that memory still rings alarms bells! LOL).

Again, my point remains - sharing any info on performance improvements, be it Vulkan engine development or of other performance optimizations, on a frequent basis - no matter how small those news would be - would go a very long way. Afterall, we the customers/users are the first to be interested in them!

On 9/17/2022 at 6:14 PM, okopanja said:

We appreciate you came forward to us, although at the moment we look like an angry mob.

As a SW developer I usually renew my personal devices at period of at least 4 years. This means I can pretty much afford to buy PC which offers the decent performance. Sizable part of customer base can afford this over prolonged periods of time, and some do it actually every year.

However, many of your customers can not afford frequent hardware updates at the frequency the professionals do, far less than developers in gaming company. This means that a number of these users are affected heavily when ever you decide to leverage additional features ( generally desired) vs Moor's law (results in higher costs).

I would say I am quite low average at owning 1 map, FC3, Mig-21, F-16 and Ka-50. Notice how little the software costs compared to the needed hardware to run this thing. I would like to point out that investing in performance is not a wasted effort, it lets your user base be more elastic in terms of hardware ranges. Better performance with aged hardware means, that these people will remain with DCS the longer time. At the same time they may spend more money on your products than on expensive hardware updates. Please consider using this arguments when encountering Sales&Marketing people.

In short: the more we spend on NVIDIA/Intel/AMD, the less we spend on your products.

Well said.  

Just to further show how complicated new engines often are.

343 Studios has abandoned their SlipSpace Engine after 7 years of development.

Please forgive my ignorance but I was researching what the switch to Vulkan means to DCS for user performance and I had the thought that with the many recent announcements of the planned maps and modules from third party developers is this a direct result of the eventual shift to Vulkan?  

With my very limited and very laymen understanding of the Vulkan API it does seem like it may be some of the increase of interest from newer developers.  I understand it may not be the primary reason (although that is my gut feeling) but at minimum Vulkan is a better option to tackle the complexities of maps and modules vs the DX12.

Anyways just curious if my train of thoughts/therory is completely wrong or kinda of on track?

Thanks

No, it's a direct result of a shift in announcement policy, made to prevent duplication of effort at the cost of possibly being a pain in the PR. They used to announce a module when it was nearly ready, now they're going to do so shortly after it's taken. As a result, a number of projects that were being developed under wraps had come out of the woodwork.

5 hours ago, SkateZilla said:

343 Studios has abandoned their SlipSpace Engine after 7 years of development.

Considering the new Engines that have come out, like Unreal Engine 5, I'm not surprised.  7 years is Ancient in computer time.  And the engine that drives DCS is multiples beyond that.

The huge problem for me is the single core CPU limit.  Your sky graphics ( especially the clouds) are very beautiful, but we're stuck on that one core, no matter what Beast of a CPU we buy.  I've been searching, and have found other simulators popping up on Modern Engines.  The best one is in pre-Alpha right now, but even if it doesn't make it, others will come in to replace it as well.  They're all running on pre-built engines like Unreal Engine 5.

You guys have to speed it up, or your going to be left in the dust.  The Standard now is Whole Earth, FPS, Vehicle, maybe even Space.  Everyone is looking for a Realistic "Arma" style sim.  This is where the majority of simmers are going to end up.

With all respect, maybe it's time you considered using one of the Modern pre-built engines, before you lose it all.

4 minutes ago, 3WA said:

Considering the new Engines that have come out, like Unreal Engine 5, I'm not surprised.  7 years is Ancient in computer time.  And the engine that drives DCS is multiples beyond that.

The huge problem for me is the single core CPU limit.  Your sky graphics ( especially the clouds) are very beautiful, but we're stuck on that one core, no matter what Beast of a CPU we buy.  I've been searching, and have found other simulators popping up on Modern Engines.  The best one is in pre-Alpha right now, but even if it doesn't make it, others will come in to replace it as well.  They're all running on pre-built engines like Unreal Engine 5.

You guys have to speed it up, or your going to be left in the dust.

With all respect, maybe it's time you considered using one of the Modern pre-built engines, before you lose it all.

Unreal 5 started it's development before Slipspace, just so you know. They might have said "a few years" but the project lead for UE5 was hired over 7 years ago.
 

Well, have you seen what that Engine can do?  I swear, they got their hands on Alien tech or something.  They're doing stuff in their demo's I never thought would be possible at this time.

On 9/17/2022 at 5:14 PM, 59th_Krieger said:

Hi all,

It's my personal opinion but getting disappointing that:

The first announcement of Vulcan happened at Nov 28, 2017.

As customer I just experience that there is no god to push enough performance out with VR from the game (without 3rd party tools). Before someone just asking I use: Ryzen 9 5900XT with ASUS 3090 OC and 64GB RAM, and yes I have spent enough time to find the best options.

If sometimes, still means that lots of already expected 'magic' still missing, like shadows generated by anything except the sun, moon and the 'tricky magic' shadows on carriers or

just a few annoying bugs:

- falling red colored snow,

- the fog still look like in the games around 2000 especially in rain,

- you hear the rain but not see it when using rain preset.

- after moved some LUAs into DLLs, the loading in menu became horribly slow,

- Mig29 cockpit 'unusable' at night since 2 years (no reaction for requests for fix especially if it has affect for ED modules)

- no fog visible in rain but rain moving into opposite direction than the wind,

- more then 90% of tracks not usable (We don't even dream that get a player where possible to forwards or backwards the time - in 2022 it is unthinkable),

- snow falling in hangar too,

and so on, many other things on 3rd party side as well.

 

At the same time 1-2 core of CPU just 'melting down' to serve everything, especially in multiplayer or missions, so sometimes some info or update would be enough to not sell the whole thousands of dollars worth equipment, and start to believe that something happening beyond the promises.

I quote again: "The first announcement of Vulcan happened at Nov 28, 2017."

 

I along with others have got to a point that we won't spend another penny on DCS until Vulkan is released. I get that Rome wasn't built in a day, but I along with many others have literally spent thousands on a gaming rig to play DCS. So to still be on 2 cores in 2022 is just beyond a joke and it's driven many away from DCS.  

I think ED pretty much knows how much DCS needs Vulkan and Multithreading.

There is no denial on that some kind of evolution needs to take place and ED has clearly stated that they follow the path of Vulkan and Multithreading to fix performance issues.

How long it takes ? Nobody knows, not even the lead coder could tell you I am afraid. I am no coder myself but an IT guy and looking at the nature of DCS it is near to assume that it is

not just complicated but yet predictable but complex, a totally different nature of a beast that every big code gets that has been carried along for many years, the 911 of flight sims,

never changed really, only got better. You turn here and bang, you fix it and break two others, together with a transition from DX11 to an OS indep. Vulkan platform, great. I am happy they

are at least working on it and not just plain let it run out as it is and say, enoug is enough. 

But It really matters how you look at it. If you are totally at it, invested a lot of time and money etc..I get it. Just stay tuned, I am sure ED delivers when it's ripe

I would not expect massive performance increase but will help for sue. 

Multicore will almost certainly deliver a massive performance boost on complex missions, particularly for people with modern CPUs, since it'll actually use the resources modern ones have. Vulkan should change less except for VR users, where it'll be a gamechanger as well, since its way of rendering VR is more optimized than what we have now (we basically render the same thing twice, Vulkan is smarter about it).

It will depend on hardware and the mission being flown, but I would fully expect a very noticeable increase.

Going from Dual Threading to Multi Threading is the Equivelent of taking your 2022 V12 out of Eco-Mode and allowing as 12 Cylinders to fire instead of just 2.

34 minutes ago, SkateZilla said:

Going from Dual Threading to Multi Threading is the Equivelent of taking your 2022 V12 out of Eco-Mode and allowing as 12 Cylinders to fire instead of just 2.

So, can you confirm, that new multithreading for DCS will run faster the more cores are used and not just go from 2 threads to e.g. 4?

2 minutes ago, Tom Kazansky said:

So, can you confirm, that new multithreading for DCS will run faster the more cores are used and not just go from 2 threads to e.g. 4?

Any announcements regarding in development features would be made by an ED Employee.

Just now, SkateZilla said:

Ok. At least you don't kill my high hopes. Thanks.