1.05 Benchmark Mission Results

[thr0tt 12]

Maybe benchmark was the wrong thing to call them but maybe stress test ? People looking at these shockingly low figures on moderately high end systems are going to be so put off.

Not sure why the 2nd bend was that much worse in performance for all, should I be getting 15fps on my low end system ? What did BIS get when they ran it and what settings as a 'goal' ?

[jasonnoguchi 11]

^... I got only 19 on mission 2 with my rig and my rig isn't exactly low end.

[Gamer 10]

Bench 1 = 122 fps

Bench 2 = 24 fps

Using latest cat drivers and ATT tool to turn off vsync

Shadows - off

graphics memory -high

Post process - disabled

AA - normal

AF - Very high

Objects detail - high

Ground detail - normal

View distance - 3600

Resolution - 1920x1200

3D Resolution - 1920x1200

Not sure what the deal is with bench 2, does not seem like it will go higher no matter what settings I use.

[PostaL 0]

The brown/yellow tree performance is a pain in the ass. And visual they look not that good either. They should remake other less perfomance hog trees. Especially in the new 1.05 campaign i have to walk in those performance hog tree areas. BI should make other trees. Trees are just as an extra to fill the island up to look it more life like. They shouldn't be the major performance impact in a game. In general you focus more on the enemy and not how ubernice the trees look whilst playing the game.

Also the barr trees(pine and such) still look pretty nice and dont have these performance problems.

I'd rather have the island full of them barr trees then run to parts of the island where i have to lower my resolution / fillrate and shadows off to be able to aim at enemies in a leafy tree area.

[neokika 62]

Test1 = 42fps

Test2 = 19fps

1680x1050/1680x1050

View Distance = 3200

Textures = Very High

Video Memory = Very High

Anistronopic Filter = Very High

Anti-Aliasing = OFF

Terrain Detail = Normal

Objects Detail = Normal

Shadows = High

Post Processing = Very High

System: E8400 4.1Ghz

2x2 DDR2 1066mhz Dominator

GTX285

[Satarix 10]

I am just thinking of buying ArmA2 and upgrading my system exclusively for it.

Could you please advise how the performance in these test1 and test2 relate to actual gameplay FPS.

I.e. if test1 = x fps and test2 = y fps then what will be average fps for:

1) campaign

2) multiplayer

A. person vs. person, say 32 ppl server (no AI)

B. multiplayer co-op (vs AI)

3) editor

I hope it is possible to derive some ratios between these 2 tests and real gameplay.

Reading this forum for 2 weeks I am very much confused as some people are happy with Athlon x2 and 9800gtx playing at very high 1680x1050 and others are not happy with i7/295. I guess the reason is some are happy with 20 fps in the editor and others need at least 40 fps online.

I think this information would be very useful for everyone considering purchasing ArmA.

[T1NG 10]

I got

Bench 1 = 22 FPS average

Bench 2 = 8 FPS average

I found this a shock because when actually playing its petty smooth and Id say runs ok, the only real time i get FPS issues is online in CTI (Player name:Martyn) after the game has been running for ages and theres a lot going on. the same would probably happen in other game types but have only played CTI so cant say. I use Athlon x2 with a 9800GTX with 2gb ram and have all the in game settings set to normal apart from the quality is set to high, thats on high because it don't let me select normal, if I do it switches back to high on its own.

I think im going to try and SLI my 9800GTX with a 9600GT which I have lying around (not even sure if thats possible), Ill report back the results I get with that.

[tonygrunt 10]

Bench2 seems to have a 25 FPS cap.

Memory default.

View Distance 2500.

1680*1050 all settings very low or disabled 25 FPS.

1900*1200 all settings very low or disabled 25 FPS.

1900*1200 all settings very high 23 FPS.

Task manager shows 1 cpu working at 90% and the other 3 cpus below 50%.

For comparison Bench1 shows some wide margins between low and high graphics:

1680*1050 all settings very low or disabled 124 FPS.

1900*1200 all settings very low or disabled 123 FPS.

1900*1200 all settings very high 50 FPS.

[zaira 10]

Maybe benchmark was the wrong thing to call them but maybe stress test ? People looking at these shockingly low figures on moderately high end systems are going to be so put off.

Not sure why the 2nd bend was that much worse in performance for all, should I be getting 15fps on my low end system ? What did BIS get when they ran it and what settings as a 'goal' ?

I would call it stress test, rather than benchark...

M8 your system isnt low end, i think it is a mid range system.

Clanmate has same cpu + 5870 and gets 14 fps, in second test.

This second test realy points that minimum fps in arma2 is cpu related.

1680x1050/1680x1050

View Distance = 1600

Textures = Very High

Video Memory = Very High

Anistronopic Filter = Normal

Anti-Aliasing = OFF

Terrain Detail = Very High

Objects Detail = Very High

Shadows = Normsl

Post Processing =OFF

B1 34fps

B2 13fps

E6600@3400 MHz 4GB Ddr2 800, 4850

You realy need i7 @ 4.2+ Ghz to play arma2 in big cities and hughe battles without laging. Sadly that is the fact. I think we will need 2 generations of cpus after i7 to play it at vsinc. Just like OF, or FSX (there is no system today that can run it at 30fps+ on large airports, with everything maxed, and it is 2006 game).

But this is greatest game there is and runs good most of the time even at my 2006 mid range PC. :)

[Guest]

Yeah I am struggling alot with my PC now, more then ever. I can still run crysis maxed easily, but Arma 2 I never, ever, have a high framerate. Usually it hovers at the 30fps mark and therabouts.

Before 1.04, i could load chernarus with an empty map and just me, it stayed at 75fps unless i went near a city. Now it doesn't go above 50fps in the same situation, but its also lower in cities too. I never had stuttering before the patch, or even on Arma 2's first build, so it has had no positive effect on me :(

[randir14 10]

The brown/yellow tree performance is a pain in the ass. And visual they look not that good either. They should remake other less perfomance hog trees. Especially in the new 1.05 campaign i have to walk in those performance hog tree areas. BI should make other trees. Trees are just as an extra to fill the island up to look it more life like. They shouldn't be the major performance impact in a game. In general you focus more on the enemy and not how ubernice the trees look whilst playing the game.

Also the barr trees(pine and such) still look pretty nice and dont have these performance problems.

I'd rather have the island full of them barr trees then run to parts of the island where i have to lower my resolution / fillrate and shadows off to be able to aim at enemies in a leafy tree area.

People have been saying this for months and I haven't seen a single BIS person acknowledge the problem. Other than AI the main performance killer is the trees, try using Kju's Proper Low Vegetation mod to see the huge FPS boost it gives without any noticeable visual degradation.

[mali robot 10]

Test 1: 33

Test 2: 16

E8400

GTX 275

4 gb ram

XP SP 3

...and a lot of programs running in the background (old video drivers, hd not defragmented).

i think that my framerate is better during normal play, can't remeber that game was unplayible, except in big cities, but beta patches improved that (haven't tried with 1.05, it shoudn't change?).

[gulag 10]

Test 1: 53

Test 2: 20

1440x900; view distance 1800; textures-normal; video mem-very high; AF-very high;

AA-normal; terrain det-normal; obj det-high; shadows-normal; PP-off.

Q9550 3.0Gb

HD 5850 1Gb

4 Gb RAM

WIN7 64 bit

[tankkiller01 0]

Test 1.. 34fps

Test 2.. 16fps

Most game settings high, aa low, obj norm

_________________________________________

Intel core 2 Quad Q9550 @3.3Ghz

BFG Tech GTX 285 OC

Asus P5N-D Mobo

4GB DDR2 Corsair XMS

Thermaltake Ruby Orb Cpu cooler

Maxtor HDD

EZ COOL 650w PSU

[Binkowski 26]

Test 1 : 20

Test 2 : 7 :(

1680x1050; 1600 VD; Everything on Normal; AA on Low; PP disabled; Shadows High

Pretty sure if I turn the shadows down and disable AA I'll get better results. Going to run that test soon and see what happens. Man I need a new video card.

[randir14 10]

I actually get better FPS with shadows on high than on lower settings, other people have had the same experience. Maybe it depends on the hardware.

[-DirTyDeeDs--Ziggy- 0]

all tests are run with 3D render @100%

settings- all V High, aa Normal, pp Low

1=34

2=18

settings- all Normal, aa Low, pp Low

1=45

2=21

settings- Lowest possible, V Low or Disabled when applicable

1=59

2=22

:cool:

Edited December 25, 2009 by [DirTyDeeDs]-Ziggy-

[Dodgeboy328 10]

1st=9

2nd=8

Almost all settings are on very low with a few disabled. The OPFOR isn't my enemy in Arma 2, my computer is. haha.

Dell XPS 1530

Intel Core 2 Duo 2.10 GHz

4.00 GB

Vista 32-bit OS

NVIDIA GeForce 8600M GT

[-SONAF-Rebel 10]

1st: 22

2nd: 14

1280*1024, AA and AF off, High-normal settings, view dist. 1222

It`s quite interesting that I do not get better results if I set the details lower.

[-bsm--sniper 10]

Hi all,

RIG in sign.

1680x1050/1680x1050

View Distance = 1600

Textures = Normal

Video Memory = Default

Anistronopic Filter = High

Anti-Aliasing = OFF

Terrain Detail = Low

Objects Detail = High

Shadows = Very High

Post Processing = Disabled

Windows 7 64b:

Benchmark 1: 32fps

Benchmark 2: 16fps

Windows XP 32b:

Benchmark 1: 59fps

Benchmark 2: 20fps

[idiosis 10]

Win7, 3.4 Quad Core with an ATI 5750, 4gb RAM and an old Audigy 2 ZS - ran this last night and was quite pleased with the results:

Benchmark 1: 45 fps average

Benchmark 2: 16 fps average

Although having said that, looking at [bsm]-sniper's comparision between XP & Windows 7 I might try dual-booting XP tonight if there's that much of difference.

[S-M 10]

Mine was running pretty sweet on xp32.

Since replacing the hard drive and installing win7 fresh, it seams a lot more "clunky" for some reason.

Installing an extra 2gb (2gb on xp to 4gb on win7) helped a little, but not by much, if i watch the system resource logs after alt-tabbing out onto the desktop, it seams to use about 50-60% of my cpu power, and the graphics card seams about the same.

Not even close to breaking a sweat, yet it still seams to chug along at 25-30fps while in game, will run the benchmark tonight

[Flash Thunder 10]

Benchmark1= 24FPS average

Benchmark2= 6FPS average

I was wondering too, what the hell is going on?

damn Chernarus and all that useless foilage!! :mad:

I actually get better FPS with shadows on high than on lower settings, other people have had the same experience. Maybe it depends on the hardware.

Yes I do also, when I dont have VERY-HIGH shadows on my FPS is 3 lower than when I have it, also Anti Aliasing is the same way I can have it on Low without any FPS loss, when with my old GPU it would always drop at least 10 frames.

This engine doesnt make sense, unless its completely our hardware. wtf??!

Edited January 25, 2010 by Flash Thunder

[Bavator 8]

Benchmark1= 24FPS average

Benchmark2= 6FPS average

I was wondering too, what the hell is going on?

damn Chernarus and all that useless foilage!!

Get Q´s PROPER vegetation tweak, 10 -15 FPS more on my rig.

Yes I do also, when I dont have VERY-HIGH shadows on my FPS is 3 lower than when I have it

AFAIK shadows on medium are calculated by the cpu, above by the gpu.

[frostybits 0]

Benchmark 1=33

2=15

VD=3200

1680x1050

v sync on

everything high

asitrop =normal

texture detail =normal

post pros =off

9550 stock

4gig ram

275gtx factory overclocked

vista 64

Edited February 4, 2010 by frostybits