Arma 3 Performance vs. Arma 2

[calin_banc 19]

I wouldn't go as far as saying "far better optimized than arma" myself, perhaps optimized well for role it plays, yes.

Frostbite 2 has it relatively easy by comparison, some fancy cover up effects, no cpu heavy AI

BF 3 is CPU heavy due to the physics in it. For multi GPU setups you'll need something with more than 4 threads (2600/2700 and so on) in order for the GPU usage to stay up and avoid some low FPS. Also, Frostbyte uses more of the CPU compared to ArmA's 2 engine.

Although the maps are smaller, the game looks and runs much better and more importantly, the control of the character and the handling/feeling of guns, how they shoot and the way they look, are far ahead of ArmA 2. There is really no point in comparing those two. Of course, a modded version of ArmA can add some nice stuff and take back some territory, but in the end is not a very smooth experience, but it has it's own strong points that still keeps us playing it.

Regarding this Alpha, personally I wouldn't expect much of an improvement of performance in the final version. Some tweaks here and there, but the overall numbers should be the same. All the games I've played so far (not much, but still) that had a pre-launch testing time, ran +/- the same in the final version. If BIS is the exception from this rule, than bravo!

Edited February 28, 2013 by calin_banc

[fx2k 1]

I'm sure they fixed the memory leak issues by now. Gamescom was almost 5 months, if that hasn't been fixed by now then we have a serious problem.

I'm sure there are going to be quite a few issues, but isn't that one of the purposes of an Alpha release? To collectively find, gather and report such?

[DShKM 18]

I'm sure there are going to be quite a few issues, but isn't that one of the purposes of an Alpha release? To collectively find, gather and report such?

I completely agree, but an issue like a memory leak should be fixed before players get to try it out otherwise, how are we supposed to test if we can't last 5 minutes in the game without crashing. Now, regarding performance, I spoke to some friends at school and online and I don't see why performance wouldn't be as good or better than ArmA 2. Even with the better graphics and new engine, nicely written code can do wonders to any game.

[sproyd 2]

I completely agree, but an issue like a memory leak should be fixed before players get to try it out otherwise, how are we supposed to test if we can't last 5 minutes in the game without crashing. Now, regarding performance, I spoke to some friends at school and online and I don't see why performance wouldn't be as good or better than ArmA 2. Even with the better graphics and new engine, nicely written code can do wonders to any game.

Ah, naive youngsters. Yes it "should" be fixed but Far Cry 2, an old AAA title from Ubisoft, features a Memory Leak even after the most current patch. And that was released in 2008. Fortunately that game used more than 2GB of RAM, even back then.

[naizarak 4]

I played through a lot of the demo scenarios in Cologne and while they were obviously decent PCs it looked and ran incredibly well compared to Arma 2. They could have been running triple SLI under there on a 3960k and the entire game in RAMDisk to fool us on what the performance would look like but I doubt that amount of cheating in the demo. Doesn't tell us how the real map will fair or what its like when the AI count goes up a lot but it was smooth 6 months ago despite the fancy graphical fidelity. My guess is we'll be pleasantly surprised with performance and the visual quality we get at the various points of performance, for the 5 minutes we get to play it between memory leak crashes.

I remember viewing the forums during E3/gamescom and I believe someone posted system specs. of the machines running the demo. I distinctly remember there being a 6-core CPU(i7 970 i think), and i believe graphics was a single gtx580. a little hazy but definitely some high-end stuff in their demo rigs. granted they were running at high settings and it was an early build

[sproyd 2]

I remember viewing the forums during E3/gamescom and I believe someone posted system specs. of the machines running the demo. I distinctly remember there being a 6-core CPU(i7 970 i think), and i believe graphics was a single gtx580. a little hazy but definitely some high-end stuff in their demo rigs. granted they were running at high settings and it was an early build

Well firstly, if it was a 6 core CPU and it was running smooth because of that then that would indicate multi-core optimisation so potentially supporting the case for improved performance, relatively, over A2?

Yes I'm pretty sure it was GTX580.

[kilroy the nerd 14]

I believe they said that if you can run ArmA 2 comfortably, you can run ArmA 3 enough. Not that it wouldn't hurt to get a better rig, of course! But like Raptor said on the first page, same engine as ArmA 2, just fancier. See cool new features, better graphics, etc.

[aammooss1122 1]

Seriously , my pc configuration is bad for now , so I just planning to have an "normal setting " hopefully my pc can undertake it ha

[DShKM 18]

Ah, naive youngsters. Yes it "should" be fixed but Far Cry 2, an old AAA title from Ubisoft, features a Memory Leak even after the most current patch. And that was released in 2008. Fortunately that game used more than 2GB of RAM, even back then.

I'll take that "naive youngster" remark as a compliment. :) I'm simply being optimistic and also assume that at this point, it would have been fixed. We could go on for ages about this, let's wait for Tuesday and find out.

Edited February 28, 2013 by ccmoreaucc

[buhrewno 10]

I was at the E3 show last year.

They were playing it on a computer with a quad core 3.2 ghz processor and a GTX 580 graphics card and the frame rate never dropped below 50

[ytfeynman 10]

I was at the E3 show last year.

They were playing it on a computer with a quad core 3.2 ghz processor and a GTX 580 graphics card and the frame rate never dropped below 50

What's the graphic settings?

[buhrewno 10]

What's the graphic settings?

I think it was on high IIRC

[sproyd 2]

I think it was on high IIRC

LOL they're not going to demo their upcoming game on Low now are they!!

Anyway, on this point I've resurrected this (2 day) old thread - new PC Gamer interview that talks about performance issues, auto-settings etc

http://www.pcgamer.com/2013/03/02/arma-3-interview-arma-3-alpha/2/

A quote

van ‘t Land: During recent weekends I’ve been evaluating the Alpha at home on a laptop, granted a fairly high-performance model—but its GPU is a bottleneck. The auto-detection worked much better than with previous Arma games. I could play comfortably without tweaking, but did go into the options to change the balance between graphical fidelity and performance as Jay pointed out.

Great news!

[Zyromkiru 10]

well I tweaked with Arma 2 and it seems Takistan has good performance for me and I can have all the goods turned on if I want and have it be completely playable with the view distance of my choice (prefer 1,500-2,000 for performance but being able to see all i need) but Chernarus seems to give much less performance and it mainly seems that object detail is the killer mainly and I find it funny how I can have certain settings on and have almost no performance drop such as vsync which usually drops more fps in other games.

I'm sure it's because of the HUGE amount of grass and trees, although even in the building areas I have better performance in Takistan than Chernarus so I'm thinking that this new game will actually be easier for me to run the game due to the type of island it is without mass amounts of trees and pure grass, etc. Plus all the time they have to optimize the game before Atlis is out.

Have noticed something I dislike though and it's t hat on multiplayer you have to change view by user profile (or whatever it is) and singleplayer you can just scroll it to your liking. Personally if you are given the choice I would love to have a scroll bar for view distance on MP although it might cause too much trouble so probably a bad idea...

[sproyd 2]

well I tweaked with Arma 2 and it seems Takistan has good performance for me and I can have all the goods turned on if I want and have it be completely playable with the view distance of my choice (prefer 1,500-2,000 for performance but being able to see all i need) but Chernarus seems to give much less performance and it mainly seems that object detail is the killer mainly and I find it funny how I can have certain settings on and have almost no performance drop such as vsync which usually drops more fps in other games.

I'm sure it's because of the HUGE amount of grass and trees, although even in the building areas I have better performance in Takistan than Chernarus so I'm thinking that this new game will actually be easier for me to run the game due to the type of island it is without mass amounts of trees and pure grass, etc. Plus all the time they have to optimize the game before Atlis is out.

Have noticed something I dislike though and it's t hat on multiplayer you have to change view by user profile (or whatever it is) and singleplayer you can just scroll it to your liking. Personally if you are given the choice I would love to have a scroll bar for view distance on MP although it might cause too much trouble so probably a bad idea...

You are not alone - I don't think anyone gets better performance in Chernarus than Takistan - I blame it on the trees, but this is a real reason so you can't complain.

I am also of the opinion that Arma 3 will generally run better than Arma 2.

I think on MP view distance is often set by the server rather than the client - please someone correct me if I'm wrong. I know in Wasteland specifically you can change it through your special in-game Player Menu to Low/Med/High.

[metalcraze 290]

Yes it is set by a server and it should stay that way. Imagine if the tank crew has a viewdistance of 2km and a dude in the enemy chopper will have it set at 4 and is 3.5km away from that tank - and all the unfair advantage it creates.

Also while Limnos isn't Chernarus do note that detail wise grass and trees have more polygons and much better textures plus clouds plus render-to-texture (which is sometimes multiple times per vehicle) - that latter stuff kills performance harder than 100 trees. Improved lighting also doesn't come for free.

Edited March 3, 2013 by metalcraze

[smookie 11]

Its possible to switch the viewdistance for seperate players by mission design. Feature has been used since OFP in PvP maps.

[Zyromkiru 10]

Yes it is set by a server and it should stay that way. Imagine if the tank crew has a viewdistance of 2km and a dude in the enemy chopper will have it set at 4 and is 3.5km away from that tank - and all the unfair advantage it creates.

Also while Atlis isn't Chernarus do note that detail wise grass and trees have more polygons and much better textures plus clouds plus render-to-texture (which is sometimes multiple times per vehicle) - that latter stuff kills performance harder than 100 trees. Improved lighting also doesn't come for free.

Although, do note that with things such as Post Processing turned on all those trees use that and will give you a big hit, also with AA turned on they like to take you down a little as well. Most everywhere else the game is COMPLETELY fine and smooth as butter. Newer things and polygon count don't bug me as it's all up to the optimization and how things are done and I have been in soo many situations with games where newer things ran better. There is at least no reason in my opinion that the game should run slower on a rig such as mine. Maybe run slower on a weak graphics card and cpu as that WILL give them a hard time.

Do keep in mind that Arma games haven't really gotten any love from Drivers yet so it seems.. If this game is popular enough it will most likely lead to better drivers meaning better performance from that as well.

[doveman 7]

I drop to 17fps or lower sometimes and it seems mainly due to the AI and a poorly optimised engine as I can get 50fps on MP servers. https://dev-heaven.net/issues/59932#note-17

I'm currently running without AA/AF or ATOC as they cause weird white dots on the trees. I find the trees/foliage looks pretty awful whatever the settings though and as others have noted, the trees cripple performance whatever the settings.

So I'm just hoping that ArmA3 is much better when it comes to processing the AI and trees.

[smiley_ie 7]

I'm gonna hang on before i buy Arma 3 have bought all Arma series totally basing this one on performance & optimization. Running good to high-end PC and Arma 2 is the only software that runs pretty bad for me no matter what settings i use.

[Leopardi 0]

I'm gonna hang on before i buy Arma 3 have bought all Arma series totally basing this one on performance & optimization. Running good to high-end PC and Arma 2 is the only software that runs pretty bad for me no matter what settings i use.

Let me guess, AMD cpu?

[ruhtraeel 1]

I remember viewing the forums during E3/gamescom and I believe someone posted system specs. of the machines running the demo. I distinctly remember there being a 6-core CPU(i7 970 i think), and i believe graphics was a single gtx580. a little hazy but definitely some high-end stuff in their demo rigs. granted they were running at high settings and it was an early build

That would be me.

Machine Used by Random BI Developer:

I am running the game with i5 2500k, GTX580 SOC and 16GB Ram and its running smoothly with stable 35+ fps (high to very high quality with AA and more;)) with little to medium headcount which at this point of game development is satisfactory i think.

Machine Used in Trailer Videos:

i7-2600, GTX 560, 8GB, SSD

Machine Used at E3:

These are the specs for the system NVIDIA kindly provided us with. Madcatz provided us with the peripherals, including the mouse (R.A.T. 3), which was cool, too.

Disclaimer: I have no affliliation with any of the hardware providers mentioned below, I'm not any particular authority on such topics, am I not being paid thousands of dollars to promote them (although.. if they're reading... :p), nor would I guarantee the game will run the same way at launch as it did on the videos. Because, of course, it'll run even better, right?!

Anecdotally, we were nicely impressed with how the game felt running; although, there was no time to do any proper analyasis of performance, and inside the presentation display cabinets it would get quite hot, which led to some instability.

Operating System: Windows 7 Enterprise 64-bit (6.1, Build 7601) Service Pack 1

Processor: Intel® Core i7 CPU 970 @ 3.20GHz (12 CPUs), ~3.2GHz

Memory: 12288MB RAM

DirectX Version: DirectX 11

Card name: NVIDIA GeForce GTX 580

Monitor Model: V3D245

Disk Model: C300-CTF DDAC256MAG SCSI Disk Device

In terms of our setup, we forgot to copy our config before the PCs were canabalized, so we're sorry about that. Generally I'd say:

Viewdistance: around 3km; varied a little between pcs

objDraw: around 2.5-2.7 km; varied a little between pcs

shadow draw: around 300m; varied.

quality settings: everything at least on high; textures very high, i think terrain quality, too.

Need to head to the airport now; hopefully that gives you guys an idea where we're at/ material to start flamewars about hardware vendors :)

Best,

RiE

[Covert_Death 11]

Let me guess, AMD cpu?

not sure why brand has to be apart of this.... lets rephrase that to "not great CPU"... i have an AMD and expect 0 problems compared to my i5 counterparts... if he has an old out dated our mid grade CPU, its going to be the same bottleneck regardless of company

[Pyrophosphate 1]

Let me guess, AMD cpu?

I've seen benchmarks where the FX 8350 absolutely crushes the 3570k at Arma 2; just about double the framerate.

[scaramoosh 10]

Its possible to switch the viewdistance for seperate players by mission design. Feature has been used since OFP in PvP maps.

View distance should be 100% at all times, it's what ruins Dayz, just makes the game so ugly.