A2 no longer benefits from 5870 crossfire

[Yapab 10]

These "built-in benchmarks" where are they to be found? I know the demo had one, but I havent seen one in the full game (DVD retail)?

ARMA mark imo is not an accurate test with ARMA2.

[node_runner 0]

These "built-in benchmarks" where are they to be found? I know the demo had one, but I havent seen one in the full game (DVD retail)?

ARMA mark imo is not an accurate test with ARMA2.

Benchmark 1 and 2 can be found in the Scenarios section within Single Player (assuming your version of A2 is patched).

What do you mean by "not an accurate test"? All of those benchmarks simply place you in the game with a fixed camera. The game engine is rendering the same objects that it would be if you were controlling a player yourself. But I don't even know what you mean by "accurate" benchmark so you'll have to please explain.

[Yapab 10]

In that excel document half the data is missing from the ARMA2 marks for various configurations.

[node_runner 0]

In that excel document half the data is missing from the ARMA2 marks for various configurations.

Correct. I said that I wasn't done yet in my post, but it is enough to get a good idea and do some comparisons. There is more than half the data there. I have yet to do 2.5 of the configurations out of a total of 12.

[Chill xl 10]

This weekend i tried the 9.12 hotfix drivers and for some reason the crossfire doesnt seem to be working altough in ccc its set. After a little research i found that 2 out my 3 cards where running at 0 mhz for both gpu and vram, only my prim. gfxcard was running at correct speeds. Ingame the result was performance levels of a single 4850.

So i changed it back again to 9.11 and the crossfire was listing all the cards correctly with the right speeds and according ingame performance levels were back aswell.

Not sure if it was just an error/bug on my client only, but its woth to check it anyway, see the ATi overdrive tab to check the speeds. Oh and one other thing, ATi cards do not like AA, when AA is enabled mostly fps is cut in half en gameplay becomes very sluggish, i use fillrate instead of AA and it works so much smoother.

Just for reference, i have an I7 920 @ 3.6ghz with 1x 4850x2 and 1x 4870 (3-way crossfire) everything is on very-high except AA and PP which are both disabled and a fillrate of 125% and the game runs mostly on 60fps, only the most harsh situations like fierce battles with lots of smokes, fires and units in the city it drops to around 45-50 fps.

[Jacky60 10]

Arma 2 does support multi gpu i'm running xfire 4890 and they work fine! Try 10.1 drivers but xfire worked since Sept. My 5970 arrives tmrw will post asap.

---------- Post added at 06:38 PM ---------- Previous post was at 06:34 PM ----------

You are clearly CPU limited-increase in quality but not in fps means CPU limitation. Graphics card is upping quality but cannot increase the volume of info from cpu!

I'm at 4.0ghz with i7 and xfire 4890 tmrw i get 5970 and suspect from this that fps will stay the same but need quality to go much higher-4/8 aa (medium or high not low AA)

[hotel 10]

Hey kklown does the in game AA actually go like this

Low 2x

Normal 4x

Hih 8x

?

I heard somewhere it was

low 1x

normal 2x

high 4 x

[kklownboy 43]

Hey kklown does the in game AA actually go like this

Low 2x

Normal 4x

Hih 8x

?

I heard somewhere it was

low 1x

normal 2x

high 4 x

Low= 2x

Normal= 4x

High=8x

Thats for ATI and Nvidia's is

- Low and medium = 2x MSAA

- High and very high = 4x MSAA

- 5 = MSAA 4x + CSAA 8x (8x called in Nvidia drivers)

- 6 = 8x MSAA (called 8xQ in Nvidia drivers)

- 7 = MSAA 4x + CSAA 16x (16x called in Nvidia drivers)

- 8 = + MSAA 8x CSAA 16x (called 16xQ in Nvidia drivers)

more or less...

The CSAA modes can be nice or just a smear/blurr depends on the game~ So 6 is better than 7 in IQ and should notice the perfomance hit there...

[Richey79 10]

Just a correction. The manufacturers states that even though you xfire more cards, you don't get access to the 2nd or 3rd card's memory bank so the memory available should be consistent, no matter how many cards are in use. Also, I don't know why.

That's right. Each GPU has to hold the textures for the whole screen in its memory, even if it's only processing a quarter of the graphics for the screen. This is because of the feedback loop that goes on between the GPU and CPU as the processor sorts out which tasks will go to which GPU.

Somebody in the thread was suggesting to turn off vsync using Catalyst. Catalyst Control Panel doesn't turn off vsync for Arma2, even when you tell it to be 'always off'. You need ATI TrayTools to be able to turn off vsync. It's worth doing, even if the Windows 7 driver signing is a pain for ATT.

[kklownboy 43]

...

Somebody in the thread was suggesting to turn off vsync using Catalyst. Catalyst Control Panel doesn't turn off vsync for Arma2, even when you tell it to be 'always off'. You need ATI TrayTools to be able to turn off vsync. It's worth doing, even if the Windows 7 driver signing is a pain for ATT.

Vysnc off with CF or CFX is prone to bad flickering. why would you want the tearing? 35fps is fine for this game

[EnD Of RoaD 10]

Have you tired -cpuCount=2

I have a Phenom Quad.. I seen a increase of about 20-30 FPS

Give it a try. Nice hardware you got there btw.

[Kubicide 0]

Hmm... I definitely see an increase with two 5870's versus one. Using 9.12HF with i7 @ 3.6GHz. ArmA2 can actually max out both GPUs at times according to GPUZ. Try using the training level for consistancy between tests. With one card ArmA2 can max out one GPU real fast; with Xfire enabled I need to crank up the AA in a thick forest for both GPUs to max out.

(Oh ya, certain effects really come alive with Xfire enabled such as blasts in the distance - very realistic fast motion in the initial blast!)

What I do notice is during MP at some servers I have very poor frames - decent ping ~70 or less but real lousy FPS. Yet other servers with similar ping are very fast and same FPS as running SP or training. Suspect there may be some sort of frame-lock or limit to keep MP gameplay 'even' when the server has issues or limited bandwidth, maybe? Something definately changes the FPS depending on the server...

[NoJoe 0]

That isn't the problem. I'm running a quad core i7 @3.8Ghz and the CPU is not stressed during the benchmark. I know how to benchmark my PC. The problem is definitely either driver related or a problem with A2, most likely driver related.

No, I think your results show pretty clearly that you are CPU limited. The fps stays roughly consistent with one, two, or three video cards within each high or low settings test, but increases when your CPU is overclocked.

There was a thread a bit ago that had dealt with the CPU "not being fully used" while running Arma, and why that is so. I'll see if I can dig it up (or, anyone else remember where that was?)

--NoJoe

EDIT:

Here's one of the threads. I can't remember if this is the one I'm thinking of or not... http://forums.bistudio.com/showthread.php?t=73537

There aren't a whole lot of conclusions in it, but the general consensus seems to be that it is an issue of scheduling: That even though there are multiple threads running simultaneously, they often depend on the results of the others, so there is a fair amount of CPU time that is spent waiting for other threads to finish.

Edited February 12, 2010 by NoJoe
added a link