We need this to be optimized.

[gallexme 1]

i get about 5-10 fps

with my

amd 6970

i7 920

but when i cycle through fullscreen /windowed fullscreen

and restarting the game often i get about 30-40fps on Ultra

so thats still bad but it works now mostly

on fullscreen mode i get

D3D Error DXGI_ERROR_DEVICE_REMOVED

can somebody help me?

[b00tsy 28]

I run it at 15 FPS and I am beyond happy with my framerate. I used to run ARMA 2 at 10!

Dito, I play currently on a very old pc that does not even meet the minimum specs to run it yet I still get around 15 fps. It is a bit to low for infantry combat, but I can't blame BIS for having a crappy pc of course. I am really just glad that it runs at all and that I can use the editor.

Having an old PC does have one advantage, you clearly see differences in average fps depending on what is placed on the map and what you are looking at. On a high end PC you wont even notice an fps difference. What I noticed the most so far is that the water in the alpha reduces the fps more now then looking at the land (because of the fish and sea weeds?), which was the other way around in arma2. In arma2 when I was looking at the sea it was the same as looking in the sky (high fps), now the sea drops 10 frames for me. I do hope they can improve the sea fps a bit more or more tweaking options to reduce quality and to remove sea vegetation and wildlife/fishes.

[AussieStig 10]

You are right, theres not game out there like arma with to much information to be processed but thats not excuse for this performance, what i mean is, whats the point of making a game that almost nobody can run on a stable frame rate? just sayin`

This doesn't really have anything to do with making a game that nobody can run, because people can run it well. I know people with much worse computers than me, running it at much better frame rates and qualities. It's just not optimized for most hardware, like it should be. Hardware like i5's and 670's seem to be optimized great, anything else feels like it's been forgotten about.

[tortuosit 485]

Hi guys, don't want to start a new thread for this. Last free FRAPS does not display the FPS in Arma III... can you recommend a different software? Nothing huge, really just want to display FPS.

Thx

[mbicko 1]

Add this to you steam parameters. -cpucount=4 - change number to your amount of CPUs. Helped my game alot.

[pablo0713 1]

Already try and its the same i try a lot of usermade methods but theres nothing user can do when we all know what is wrong here

[white 1]

Hi guys, don't want to start a new thread for this. Last free FRAPS does not display the FPS in Arma III... can you recommend a different software? Nothing huge, really just want to display FPS.

Thx

i use the free msi afterburner for gpu and coretemp for cpu (since it measures independent per core temp).

ALSO

ive ran a couple of tests, i posted it here:

http://forums.bistudio.com/showthread.php?147533-Low-CPU-utilization-amp-Low-FPS&p=2319002&viewfull=1#post2319002

[Threedog1421 1]

So what are you saying is we shouldn't be mentioning abysmal performance issues just because it's Alpha and is not a time for optimizations? ...

What about other bugs, following your logic we shouldn't be mentioning about them either? I mean it's Alpha after all and things are meant to be bugged / unfinished.

Yeah...

Tapatalked.

That's exactly what we're saying. So sad my 1st post has to be a semi-slam thread. People that are crying about optimization have no idea what there talking about and shouldn't even making these threads. The state of the game is exacly where BI wants it to be. People like the OP need to understand this version is no where near in any sense what the completed version will be.

[Minoza 11]

^^Actually you have no idea what you're talking about. If needed I'll explain why, but I already did across several performance topics. Oh and let me guess, Arma 3 is your first Arma game...

[Threedog1421 1]

^^Actually you have no idea what you're talking about. If needed I'll explain why, but I already did across several performance topics. Oh and let me guess, Arma 3 is your first Arma game...

Yes it is, but I've Alpha'd plenty of games to know what I'm talking about. People shouldn't be expecting a polished product at this point. If there's a issue with the engine that's different. No one should be expecting finished performance is all I'm saying. And that's a fact. Alpha is a Alpha, not meant to be looked at as a finished game. To many people these days get into alphas for the wrong reason. They want use to report game play bugs. Not perfomance issues...

[caldrin 1]

I will have to check what FPS the game is running at but it seems really smooth when im playing it.

I have everything maxed out, but have turned post processing right down as i dont like the effects..

Intel i7 2700k @ 4.5ghz

8GB DDR3

Geforce 670 2GB OC from factory.

SSD for OS

7200rpm drive for games and stuff

Win 7 x64

Edited March 7, 2013 by caldrin

[drakedaeron 13]

I turn with :

Display mode : Full Screen

Resolution : 1600x900

Aspect Ratio : 16:9 - Wide

Interface size : Large

Visibility : 1613

Object : 1111

Shadow : 100

Brightness : 1.0

Gamma : 1.0

Rendering Res. : 1600x900 (100%)

VSYNC : Disabled

Antialiasing : Disabled

PPAA : SMAA High

ATOC : Disabled

PostProcesses quality : Disabled

HDR Quality : Low

Anistropic Filtering : Ultra

Pip : Low

Dynamic lights : Low

Texture Quality : High

Objects Quality : Low

Terrain Quality : Standard

Cloud Quality : Disabled

Shadow Quality : Disabled

Particles Quality : Standard

With my GT540M 2go, i7 2670QM 2.2GHZ, 6GO RAM.

My CPU is used in 30%, and I have 30/40FPS with no action, 30FPS in forest, 25/35 in Combat IN EDITOR

And in multiplayer 25/28FPS.

I am very satisfied, but my CPU is not used a lot, me waits to see if Bohemia gets better optimized the game, I could can be credit note of better performance, but I think that I am on the verge of the possibilities with my PC.

[Desrat 0]

35-40fps on my rig, which is more or less the same as A2 ran, so I'm not complaining

Core i5 2500k

Z78 chipset mainboard by Gigabyte

8GB DDR3 1600mhz

Radeon HD 5830

120GB Agility 3 SSD

Win7 X64

[antoineflemming 14]

Yes it is, but I've Alpha'd plenty of games to know what I'm talking about. People shouldn't be expecting a polished product at this point. If there's a issue with the engine that's different. No one should be expecting finished performance is all I'm saying. And that's a fact. Alpha is a Alpha, not meant to be looked at as a finished game. To many people these days get into alphas for the wrong reason. They want use to report game play bugs. Not perfomance issues...

They want you to test their game and report issues, period. If there's a performance issue that is caused by their game, then they'd want you to report it. This isn't some FPS Alpha where all you are testing is multiplayer connectivity and whether or not you jump high enough. I've alpha'd and beta'd plenty of games too. Performance issues are definitely something to report. No one is expecting a polished product. But people do expect you to have fixed issues that were present in previous games. But seeing as you just joined today, and no next to nothing about the ArmA series or community, I'd agree with Minoza. You don't know what you're talking about. Performance is a priority. All the new features and graphics don't mean a thing if a player can't actually run the game. And, contrary to what is said by all of you who do run the game great, it's NOT always the user's hardware that is the problem.

[spamurai 3]

Performance for me, is about stability.

I've heard all kinds of people post FPS numbers telling others what should be "acceptable". For me, the day I first turn on a new ArmA game I go to the editor.

I run around the open world and watch how "stable" the FPS is. I look for the variance in it's max and min levels and try and see what things appear to affect the FPS the most. I then start tweaking video settings to try and find a happy medium where the Max and Min are as close together (stable) as possible. Surprisingly this is much harder to do in ArmA then any other game. Things that should help improve FPS have little to no effect, things that have an effect introduce new problems to consider. Finding that bottleneck is tough.

My current system:

Intel 3.2GHZ i7-3930K (6 Core) LGA2011

16GB 1866mhz RAM

MSi Power Edition GTX670 (2GB)

AuzenTech X-Meridian 7.1 2G

Win7 (installed on SSD)

ArmA 3 (installed on SSD)

My hardware setup is no slouch. I'm lucky in having a system that exceeds the specs many ArmA players get to play A3 on. It easily plays any game or application I've currently encountered with exceptional performance. Could it be more powerful? Sure, I could add more memory into it... increase the size of the video card.... the CPU is already among the largest available, but the question is should I have too? I don't think so. I also figure that adding even more RAM and a larger video card would only have meager effect because of where the bottlenecks are in the Alpha build.

So that leaves us with working inside the constraints of the ArmA settings...

Now, what should I expect for a basic target benchmark? Of course, everyone has given values all over the board as to what is "acceptable", and there are also those people who insist that FPS doesn't matter in ArmA. For me, FPS does matter and playing any game at less then 20FPS has a perceptible effect on me. It's uncomfortable in that things feel de-sync'd. It's hard on the eyes, not to mention tough on basic input/output controls and reflexes in the game.

So, it does matter. But I'm realistic. I know that I can expect +60FPS in most of my games, but I should be ok in ArmA with less as long as it meets a standard.

The standard I'd like to see is 30FPS at the lowest, most overloaded scenario's of the game. If the game averages between 40-50 (or higher) FPS in typical situations then all the better.

Currently in the Alpha, the lowest value routinely encountered is 19FPS and the highest is +60 FPS. So... currently, the experience at any given moment is anything between "unacceptable" to "more then acceptable". How I evaluate this is less in focusing on how low the FPS gets and more on the fact that the FPS fluctuates almost 40FPS up and down the scale... often while standing in place and simply panning in a circle. And this is in the editor with an empty island... wait until we add a mission with more activity in it and an actual game to play with bullets and explosions flying around.

That's not "stable".

Stability is what we want. It is one thing to enter an area with more complex visuals like near the airfield and have the FPS drop, but remain stable at a lower then average value.... but to have the FPS constantly racing up and down creates image problems.

When presented with this situation in game I stop, and enter the settings and start experimenting with switches and controls to see what has the greatest influence on performance. How can I try and stabilize the frame rate to a more acceptable value. What puzzles me more often then not in ArmA, is that many times... it doesn't matter what setting I adjust, the FPS changes an insignificant amount, 2-5FPS sometimes. Turn things to Ultra... set them to Low, the change is often imperceptible. Then, when there is a setting that has a notable affect, it often introduces an unwanted image problem as a tradeoff. It's frustrating to have to make concessions that are often no better then just settling for a less then ideal state.

Going back in time.... this challenge has existed from the very first game I played of ArmA 1. Now, Sahrani island wasn't as visually complex as Charnarus or Stratus, but it initially introduced all the performance challenges that can be expected of an ArmA game. Many years ago, my ArmA 1 computer was a mere fraction of the hardware I have today, ArmA 1 ran with low FPS, and there plenty of image problems like LOD popping and Model flicker. It was tough; it was unacceptable, and it was something I successfully worked at eliminating from my ArmA experience. First, I worked at tweaking my ArmA settings as best as I could. Reducing visuals where ever possible. Then I upgraded hardware... first with a better Video Card and when that didn't work; I mounted a better, faster HDD... and then finally upgrading the CPU to a full quad core processor. The happy ending to the story is that all the time and effort spent improving things paid off. ArmA 1 ran like a champ. All details set on high, no image problems and most importantly no compromises in performance. Nothing stood in the way of playing the game. Performance was never given a second thought.... it was just the feeling of being immersed in the atmosphere and gameplay of ArmA. Absolute fun.

Then came ArmA 2

With the new game, the promise of a more visually complex world to explore. All the things we wanted for Sahrani Island... and of course with more complexity came the expectation that the game would run with a lower level of performance. That was expected, but was not expected was a return to the early days of those initial games of ArmA 1 and all of it's challenges. That was depressing. All the new "fluff" added to the game, better grass shaders, more varied forests, bigger map... all of which should enhance the atmosphere had now set everything back to ground zero. Time to start all over again.

Now, eventually ArmA 2 improved in various ways until it's performance was very good.... but always with a compromise somewhere and it never reached that high plateau of ArmA 1 where you just didn't think about it any more. Maybe that is why I never really got into ArmA 2 like I did Armed Assault and didn't play it as much.

Fast Forward to March 5th, 2013 and the ArmA 3 Alpha. Here we go again! Only this time, I think we are in much better, but not perfect shape. However, given the state of the game on opening day of the Alpha and how much time we have for a Q3 release... I think we have a great potential to bring back a quality of ArmA gaming on release day that I have not experienced since the last time I was cruising around Sahrani in Armed Assault.

Edited March 7, 2013 by Spamurai

[Threedog1421 1]

They want you to test their game and report issues, period. If there's a performance issue that is caused by their game, then they'd want you to report it. This isn't some FPS Alpha where all you are testing is multiplayer connectivity and whether or not you jump high enough. I've alpha'd and beta'd plenty of games too. Performance issues are definitely something to report. No one is expecting a polished product. But people do expect you to have fixed issues that were present in previous games. But seeing as you just joined today, and no next to nothing about the ArmA series or community, I'd agree with Minoza. You don't know what you're talking about. Performance is a priority. All the new features and graphics don't mean a thing if a player can't actually run the game. And, contrary to what is said by all of you who do run the game great, it's NOT always the user's hardware that is the problem.

Sorry bro I know exactly what I'm talking about. The problem here is that more than anything the new community of players are Day Z crossovers who got it blah blah blah. They played BF3,CoD and the rest. I fall into that group except I never jumped on the DayZ bandwagon. I joined becasue I want something more than the status quo. If someone has a issue they can't run the game on a recent gen system then there is definitely something wrong in there system that has ZERO to do with the game. Was the same in BF3 and other games. People were blaming the game when odds are they're working with a terribly old OS install,virus infested etc system. System not updated etc...You see where I'm going. I'll ignore your jab at me being a newb to the series, 20 years in IT tells me otherwise. I've done enough research on the game and yes it's a HUGE CPU intensive game. I won't compare this to BF3 or any other game, even I know it's a totally different beast that anything out there. My point is that all these "Optimize this already" has zero plaze in the development. Dev's aren't going to be optimizing till later in the Beta stages. THat's how every game is. And yes I'm not sure how many Alpha/Beta Builds they will be running. It'll be fine the end no worries here...

[Devil_Inside 10]

Sorry bro I know exactly what I'm talking about. The problem here is that more than anything the new community of players are Day Z crossovers who got it blah blah blah. They played BF3,CoD and the rest. I fall into that group except I never jumped on the DayZ bandwagon. I joined becasue I want something more than the status quo. If someone has a issue they can't run the game on a recent gen system then there is definitely something wrong in there system that has ZERO to do with the game. Was the same in BF3 and other games. People were blaming the game when odds are they're working with a terribly old OS install,virus infested etc system. System not updated etc...You see where I'm going. I'll ignore your jab at me being a newb to the series, 20 years in IT tells me otherwise. I've done enough research on the game and yes it's a HUGE CPU intensive game. I won't compare this to BF3 or any other game, even I know it's a totally different beast that anything out there. My point is that all these "Optimize this already" has zero plaze in the development. Dev's aren't going to be optimizing till later in the Beta stages. THat's how every game is. And yes I'm not sure how many Alpha/Beta Builds they will be running. It'll be fine the end no worries here...

lol. By the way you talk, you actually sound like a 16yo who actually did jump on the DayZ bandwagon, and that's exactly where you're coming from. If you actually had 20 years in IT, you wouldn't be making those BOLD statements about how it will be a totally different game by the time it gets released. This game has a long history. And all games in the series are plagued by mostly the same issues, which are hardly fixed from one release to the next. There might be some room for performance improvements, but it's not as huge as you're implying. Otherwise most of the optimization would be already done.

[white 1]

There might be some room for performance improvements, but it's not as huge as you're implying. Otherwise most of the optimization would be already done.

the game only uses 2 cores, i see a huge possibility for performance improvement here. and since this is an alpha i hope thats possible to address. sadly bohemia hasnt recognized this yet as true. hope they do.

i have a true 6 core (amd phenom 2 x6 thuban) and the usage doesnt go higher than 35%, meaning it only uses 2 cores. same happens for people with 4 cores, they only get 50% usage.

like ive said in another topics, yes windows shows usage between all cores but never higher than 35% on all combined, that happens because it switches the 2 cores its using at any given moment between the avaiable cores, that happens fast enough to make the unaware think that all cores are working. thats just windows doing its job and evening out the workload.

and thats proven by switching off 4 cores and getting the exact same performance with higher cpu usage.

[Pyrophosphate 1]

CPU usage as a percent doesn't mean what most people think it means, and as reported by Windows, doesn't mean much of anything at all. The game detects how many cores it has to work with and balances the workload accordingly.

Windows moving an entire running thread from one core to another is not Windows doing its job. That is a performance issue. It slows things down.

Name any complex application or game that reliably uses multiple cores at anywhere near 100% usage. Prime95 is not a complex application, neither is Bitcoin mining. Neither is any kind of benchmark. If something is using multiple cores at near 100% usage, it's a program designed specifically to do that.

[antoineflemming 14]

CPU usage as a percent doesn't mean what most people think it means, and as reported by Windows, doesn't mean much of anything at all. The game detects how many cores it has to work with and balances the workload accordingly.

Windows moving an entire running thread from one core to another is not Windows doing its job. That is a performance issue. It slows things down.

Name any complex application or game that reliably uses multiple cores at anywhere near 100% usage. Prime95 is not a complex application, neither is Bitcoin mining. Neither is any kind of benchmark. If something is using multiple cores at near 100% usage, it's a program designed specifically to do that.

I'm pretty sure neither ArmA2 nor ArmA3 balance the workload across all cores. At least, it doesn't for my system.

[white 1]

CPU usage as a percent doesn't mean what most people think it means, and as reported by Windows, doesn't mean much of anything at all. The game detects how many cores it has to work with and balances the workload accordingly.

Windows moving an entire running thread from one core to another is not Windows doing its job. That is a performance issue. It slows things down.

Name any complex application or game that reliably uses multiple cores at anywhere near 100% usage. Prime95 is not a complex application, neither is Bitcoin mining. Neither is any kind of benchmark. If something is using multiple cores at near 100% usage, it's a program designed specifically to do that.

your point becomes moot when disabling 4 of my six cores gives out the exact same performance with higher cpu usage.

and the average % of usage is directly linked to how many cores one has: 2 cores, 80%+ usage, 4 cores 50% usage, 6 cores, 35% usage. the hard drivers limits it you say? well soem of us have high end ssd´s. the memory limits you say? i say 16gb otta be enough. the gpu limits you say? well my 660ti oc on ultra gets 45% usage with vsync off no matter the settings.

all evidences point to the same thing, this game is dual core and thats not enough.

Edited March 8, 2013 by white

[Ch3v4l13r 1]

@Op: I think wanting 60 fps stable isn't realistic and has never been so when a ARMA game came out. A reasonable fps to expect with a mid-high end system is a stable fps between 30-50 fps(With stable i mean during slow and intense moments, not while your just looking at a plain field where nothing is happening) with setting between standard and ultra. Right now i am getting 30-50 with some huge unexplained drops to 5-6 from time to time. Apart from those random frame drops the game seems to be running better than i would have expected, but there is certainly room for improvement and those extreme frame drops need to be ironed out.

System:

i5 2500k @ 3.6

8GB RAM

GTX570oc

Win7

Edited March 8, 2013 by Ch3v4l13r

[metalcraze 290]

I've installed Chernarus under ArmA3 and I'm losing ~20 fps on worse settings compared to ArmA2 (30 fps at 2.5 km VD as opposed to 50 at 3.2 km VD in A2... and that's with clouds being disabled and HDR set to low) - so there's definitely something off with ArmA3.

[Pyrophosphate 1]

your point becomes moot when disabling 4 of my six cores gives out the exact same performance with higher cpu usage.

and the average % of usage is directly linked to how many cores one has: 2 cores, 80%+ usage, 4 cores 50% usage, 6 cores, 35% usage. the hard drivers limits it you say? well soem of us have high end ssd´s. the memory limits you say? i say 16gb otta be enough. the gpu limits you say? well my 660ti oc on ultra gets 45% usage with vsync off no matter the settings.

all evidences point to the same thing, this game is dual core and thats not enough.

I don't think you understand computer architecture. There's more to memory than how much you have. No memory in the world can keep up with your CPU. No SSD in the world can even almost keep up with your CPU. When your CPU requests some piece of data from memory, if it doesn't immediately have something else already there to do, it has to wait ~15-30 CPU cycles to actually get that memory. If the game tries to interface with the memory too much, that will both slow it down and reduce the apparent CPU usage, which happens to be what people are saying is the problem.

[Hellhound 10]

Name any complex application or game that reliably uses multiple cores at anywhere near 100% usage. Prime95 is not a complex application, neither is Bitcoin mining. Neither is any kind of benchmark. If something is using multiple cores at near 100% usage, it's a program designed specifically to do that.

Any 3D rendering program will do that for you (100% CPU usage on all cores). And these programs are not designed as benchmark applications, but to render images/movies (and therefore take as much advantage of your CPU as it can). When i render with Lightwave 3D it uses all my 8 HT cores of my core i7 CPU to the full extent, and i'm sure it's the same with Maya. 3Dmax, XSI, Houdini etc..).

Edited March 8, 2013 by Hellhound