Arma 3 being upgrade to 64 bit

[Frankdatank1218 39]

I am not versed in engine coding at all, but is it even possible for BIS to, for instance, upgrade Arma 3 as is to use 64 bit or additional cores? 

I'm getting sick of the constant poor performance and I just want to know, is this something we have to wait for in Arma 4? Or is it possible to make 32bit application into 64 bit or add more core usage into the game?

[austin_medic 109]

Nope. Far to many barriers to climb over at this point in the game, especially considering the RV engine is over 15 years old, and it was literally built to function with a single thread, later on having multi-threading added on top of it to some stuff thats close to the top of the pile of code (e.g AFM, PhysX). Modifying anything on the bottom of the stack would likely break many things that depend on it functioning in that specific way, and BIS doesn't have the resources to keep their company floating while they go build the engine, Enfusion is supposedly a bit better since they did some tearing down of low level systems and rebuilt them to better utilize modern computer technology, but that for arma is a long long way off. There is no simple click a button and it magically becomes 64 bit.

[Frankdatank1218 39]

What if they finished Enfusion, and ported the content in?

Other games, mainly MMO's, have transferred to new engines. Isn't what they plan to do with Dayz anyway? Drop it into Enfusion when it is done?

For instance, Insurgency is transferring from Source to UE4.

Isn't it possible, to instead of add additional mutli-threading into RV4, just drop the assets and scenario scripting, and gameplay mechanics into Enfusion? 

I mean, many pieces of content from Arma 2, with some tweaking entire campaigns, have been ported into Arma 3 with a new engine version. For goodness sake with minimal tweaking you can run Dayz's map in A3, just with issues branching from small engine differences like no proper building destrcution in Dayz causing weirdness, but with the Dev's level of control and toolbox shouldn't that kind of porting be possible with proper optimization? Since Enfusion is still very much an RV relative, wouldn't the professional makers at BIS be able to port in such a way?

[Deadfast 43]

Keep in mind that the application's bit-ness has nothing to do with threading at all. The one critical problem with a 32bit application is that it can only address (keep track of) up to 4GB of memory. Arma gets around this limitation by only loading the content it really needs into memory ("streaming").

 

This is of course bottlenecked by the speed at which data can be read from the hard drive, hence why SSD is recommended (it can load data much faster than a regular magnetic drive). Without an SSD you may notice the game briefly freezes every once in a while as you travel through the terrain since new assets are being loaded.

 

A 64bit exe could pretty much do away with this problem as you could load assets from a wider radius around the camera.

 

What a 64bit exe will not fix, however, is your average FPS. If you are standing still and yet your FPS is low then problem is elsewhere. If there aren't a lot of AI moving around then the graphics card is the problem. If you do have a lot of AI or stuff happening in general then the problem is the CPU, an area where Arma could certainly do with some optimization, both by using better parallelism and by making the code more efficient.

 

 

Would it be possible to make Arma a 64bit application? Certainly, VBS has done it and it originates from the same engine. Would it help? Probably not much.

[chortles 263]

What if they finished Enfusion, and ported the content in?

Nope.

Other games, mainly MMO's, have transferred to new engines. Isn't what they plan to do with Dayz anyway? Drop it into Enfusion when it is done?

Supposedly what's available internally (in non-public builds) is already considered Enfusion.

Isn't it possible, to instead of add additional mutli-threading into RV4, just drop the assets and scenario scripting, and gameplay mechanics into Enfusion?

Not if certain previously-stated ambitions are fulfilled, i.e. switching from P3D (basically Arma/DayZ-exclusive) to FBX (an industry standard) and deprecating SQF as much as possible if not 100%, in favor of EnScript which came from Take On Mars.

[DAP 619]

BIS doesn't have the resources to keep their company floating while they go build the engine.

Oh, yeah.. but enough for buying T-72 tank, yep.

[Greenfist 1863]

Oh, yeah.. but enough for buying T-72 tank, yep.

A new engine would set them back much more than that $50,000.

[DAP 619]

A new engine would set them back much more than that $50,000.

Yep. But + $50K or - $50K? Yeah? And don't forget, how much MANW cost. Not enough resources? Don't make me laugh. They have enough resources. But uses them in another project. Yeah. Zombies. 

[chortles 263]

*cough* And in netting themselves some new devs... and opening an Amsterdam office... and allocating resources for another non-Arma project (can you say original IP VR?)...

[teabagginpeople 398]

Oh, yeah.. but enough for buying T-72 tank, yep.

You leave frank the tank out of this. he helps the devs gets all the bitches. He was even ordained a priest to preside over one of the devs weddings.

[Mirek 166]

I dont think anyone here have any right, to comment on how anyone should spent his own money. Be it buying a tank, or investing in new projects. 

[DAP 619]

I dont think anyone here have any right, to comment on how anyone should spent his own money. Be it buying a tank, or investing in new projects. 

So let's close this forum. Feedback tracker is enough. Actually... we talking about Bohemia's abilities and resources for making new engine for Arma. I think, that they have those abilities and resources. Why they don't do that - another question. And yes, current engine is pretty old and needs replace. And yes, i don't think that they will make new engine for Arma. Another priorities. 

[Greenfist 1863]

So let's close this forum. Feedback tracker is enough. Actually... we talking about Bohemia's abilities and resources for making new engine for Arma. I think, that they have those abilities and resources. Why they don't do that - another question. And yes, current engine is pretty old and needs replace. And yes, i don't think that they will make new engine for Arma. Another priorities. 

They are making a new engine. Just not for Arma 3. And why would they replace the engine of a 3 year old game? They couldn't possibly make their money back from it.

[DAP 619]

They are making a new engine. Just not for Arma 3.

 

new engine for Arma.

 

why would they replace the engine of a 3 year old game?

I'm not talking about Arma 3. I'm talking about Arma series. 15 years on same engine with same bugs and "engine-blocks". And this damned config structure. Again and again. And I sure, that Arma 4 will be on same engine again. Bohemia. Bohemia never changes.

[Frankdatank1218 39]

Keep in mind that the application's bit-ness has nothing to do with threading at all. The one critical problem with a 32bit application is that it can only address (keep track of) up to 4GB of memory. Arma gets around this limitation by only loading the content it really needs into memory ("streaming").

 

This is of course bottlenecked by the speed at which data can be read from the hard drive, hence why SSD is recommended (it can load data much faster than a regular magnetic drive). Without an SSD you may notice the game briefly freezes every once in a while as you travel through the terrain since new assets are being loaded.

 

A 64bit exe could pretty much do away with this problem as you could load assets from a wider radius around the camera.

 

What a 64bit exe will not fix, however, is your average FPS. If you are standing still and yet your FPS is low then problem is elsewhere. If there aren't a lot of AI moving around then the graphics card is the problem. If you do have a lot of AI or stuff happening in general then the problem is the CPU, an area where Arma could certainly do with some optimization, both by using better parallelism and by making the code more efficient.

 

 

Would it be possible to make Arma a 64bit application? Certainly, VBS has done it and it originates from the same engine. Would it help? Probably not much.

My CPU is fine. It's an i7 clocked to 4.0 Ghz. It just dips alot, there is no rig on earth that can reliably crank Arma 3 at 60fps in busy urban battles, I made a benchmark mission with about 50 units in Kavala, it never goes above 45 fps, it is usually in the 30's, until the battle ends and the AI isn't doing much thinking anymore.

Why? 

Because the game barely uses other cores and the main core is where all the AI calculation goes, so my much stronger CPU works much worse than it could. 

[IndeedPete 1038]

My CPU is fine. It's an i7 clocked to 4.0 Ghz. It just dips alot, there is no rig on earth that can reliably crank Arma 3 at 60fps in busy urban battles, I made a benchmark mission with about 50 units in Kavala, it never goes above 45 fps, it is usually in the 30's, until the battle ends and the AI isn't doing much thinking anymore.

Why? 

Because the game barely uses other cores and the main core is where all the AI calculation goes, so my much stronger CPU works much worse than it could. 

 

Keep in mind that you cannot "just parallelise" things. Even if you (hypothetically) load every single calculation into its own thread, it's not guaranteed that the end result is much faster. Some thread probably requires caclulations from other threads and has to wait for them to finish. Some threads need to be synchronised properly so that, for example, the image on your screen matches with the sounds being played, movement of the objects etc. And what about resources that have to be shared by different threads? And proper allocation, scheduling, and synchronisation needs resources as well. I'm sure there's potential for optimisation but we might just have to accept that a game like this will never run in 60+ FPS.

[calin_banc 19]

 

And proper allocation, scheduling, and synchronisation needs resources as well. I'm sure there's potential for optimisation but we might just have to accept that a game like this will never run in 60+ FPS.

 

Correct, ArmA 3 most likely never, others, yes. Most definitely. It's not the game that's too complex, it's the tech behind it that it's old and inefficient.

[domokun 515]

Correct, ArmA 3 most likely never, others, yes. Most definitely. It's not the game that's too complex, it's the tech behind it that it's old and inefficient.

Others? Apart from AoS, do you know many other franchises out there that multi-thread well?

Everyone agrees that the RV engine is old and inefficient. Even BI publicly says so. So what's new?

They're working on a new engine, Enfusion, which should alleviate some of these bottlenecks but its proving more difficult than anticipated.

Maybe BI could visit the AoS devs and gain some insight into multi-threading Enfusion.

In the meantime, if you have a magic wand, by all means, please wave it!

[rekrul 7]

Ugh, again with the 15 year old engine arguments, not to mention 32 vs 64 bit.

 

Well, according to your logic, this was made on an 18 year old engine...

 

Pretty amazing, huh?

[xxbbcc 6]

Keep in mind that you cannot "just parallelise" things. Even if you (hypothetically) load every single calculation into its own thread, it's not guaranteed that the end result is much faster. Some thread probably requires caclulations from other threads and has to wait for them to finish. Some threads need to be synchronised properly so that, for example, the image on your screen matches with the sounds being played, movement of the objects etc. And what about resources that have to be shared by different threads? And proper allocation, scheduling, and synchronisation needs resources as well. I'm sure there's potential for optimisation but we might just have to accept that a game like this will never run in 60+ FPS.

 

I'd like to highlight this to everyone who thinks that doing everything in parallel is easy. It's very, very hard to get it right and it's even harder to properly test. I'm pretty sure that if it was easy, it would've been done by Bohemia long time ago. It's in their best interest to have the game run as well as possible on as many computers as possible - it simply means fewer bug tickets, fewer annoyed customers and more sales.

[calin_banc 19]

Others? Apart from AoS, do you know many other franchises out there that multi-thread well?

Everyone agrees that the RV engine is old and inefficient. Even BI publicly says so. So what's new?

They're working on a new engine, Enfusion, which should alleviate some of these bottlenecks but its proving more difficult than anticipated.

Maybe BI could visit the AoS devs and gain some insight into multi-threading Enfusion.

In the meantime, if you have a magic wand, by all means, please wave it!

 

Frostbyte, Cry Engine, RAGE (the engine behind GTA), Snowdrop (The Division), whatever they're using for The Witcher 3, probably what's behind Quantum Break, etc. They've sold 3 million units of Day Z until the beginning of 2015.

 

Anyway, according to Steam Spy, Day Z sold about 3.4 mil and ArmA 3 2.3 mil so far. Most likely they have the resources to pull it of ;).

 

 

I'd like to highlight this to everyone who thinks that doing everything in parallel is easy. It's very, very hard to get it right and it's even harder to properly test. I'm pretty sure that if it was easy, it would've been done by Bohemia long time ago. It's in their best interest to have the game run as well as possible on as many computers as possible - it simply means fewer bug tickets, fewer annoyed customers and more sales.

 

Other did it, it's possible.

[I give up 152]

Frostbyte, Cry Engine, RAGE (the engine behind GTA), Snowdrop (The Division), whatever they're using for The Witcher 3, probably what's behind Quantum Break, etc. They've sold 3 million units of Day Z by the beginning of 2015.

 

Anyway, according to Steam Spy, Day Z sold about 3.4 mil and ArmA 3 2.3 mil so far. Most likely they have the resources to pull it of ;).

 

 

 

Other did it, it's possible.

Thats how we get it. All the rest are fairy tales.

https://forums.bistudio.com/topic/188154-ram-management-serious-question/page-3#entry2999424

[CaptainAzimuth 714]

Frostbyte, Cry Engine, RAGE (the engine behind GTA), Snowdrop (The Division), whatever they're using for The Witcher 3, probably what's behind Quantum Break, etc. They've sold 3 million units of Day Z until the beginning of 2015.

Anyway, according to Steam Spy, Day Z sold about 3.4 mil and ArmA 3 2.3 mil so far. Most likely they have the resources to pull it of ;).

Other did it, it's possible.

You can leave Snow Drop (The Division Engine) out of it. It was tailored from the ground up specifically for the designing of that game.

[calin_banc 19]

Each engine is tailored for the game you're making, exposing or adding features as required or heavily modifying it, like Cry Engine was for Star Citizen. Even so, "tailored from the ground up specifically for the designing of that game" is what this game/series requires it as well and what everyone was talking about for quite some time. Besides, Bohemia is using their engine in more than one game.

[DAP 619]

Well, according to your logic, this was made on an 18 year old engine...

Pretty amazing, huh?

Please, don't make me laugh. Try check changes in each version of UE and check changes in each version of RV, yeah? PhysX, Sound Engine, Particle Effects... Again. Don't make me laugh. And yes, according to this logic - OTHER engines make big step forward. But not RV. RV even not moved to x64. But we got a plancton module. Brilliant, I think. Just brilliant.