LOW FPS USING GTX 570 - help pls!

[magnus28 1]

Hi,

No matter what, when playing combined ops, my FPS usually drops to around 14-20 with textures on high, AA on x4, and HDR on very high.

If I put everything to normal ansiotropic, aa, object and texture details, etc. I usually can get my fps to average 20-24.

If I put 3D resolution to 100% (1920x1080), as well as resolution (1920x1080),

and HDR to normal, It usually bumps up my fps by around 10, but then after a minute it usually tapers down to about a +5 gain.

(oh by the way this is playing Chernarus warfare)

If I play a takistan map, I can usually jack up 3d resolution to 200%, HDR on very high, object detail very high, and AA to 4x.

Now I know there are mods out there that reduce vegetation, low grass etc, But i'm looking for a fix without adding mods so I can play on any server.

My Specs.

AMD 1090T OC -> 4.1 Ghz.

ASUS M4A8TD PRO

EVGA GTX 570 OC -> 800/1600/2000

G.Skill PIS 160000 4GB 2000 Mhz @ 9-9-9-24

Standard 1.5 GB HD @ 7900 rpm

Monitor is SONY 40" 60 Hz @ 1920x1080 (note no fps difference than my Benq 24")

So to summarize,

-Chernarus Warfare, or Zargabad Life, have to turn almost everything down to normal to achieve 20-30 FPS

- Takistan maps, everything mostly High, get about 25-30 FPS.

**one final note**

There are several settings that I will not change - such as view distance.

I usually set it between 3000-4000 in the team menu. This is because I fly alot, and need to spot ground targets from afar.

The grid level I usually put down to 1.

I also set the view terrain in the graphics menu down to 500 (which doesn't affect the view distance in game).

please help boost fps, as with my system, I should be getting way higher than 20+fps!!!

[burdy 11]

http://forums.bistudio.com/showthread.php?t=127236 -- Post in my thread, same exact issue with a 470, http://dev-heaven.net/issues/26297 and heres the Dev note, vote it up

[PuFu 4600]

what is your ATOC value used? what are the drivers you are using?

[magnus28 1]

don't know what ATOC value is, drivers are all up to date.

I've tried using the latest beta version from Nvidia, but it seemed to make it worse, so I rolled back to latest WHQL certified drivers for windows 7

[folgore_airborne 1]

ATOC value can be seen in the Arma2 profile file. If you have Win7 look into user\documents\arma2.

[PuFu 4600]

don't know what ATOC value is, drivers are all up to date.

I've tried using the latest beta version from Nvidia, but it seemed to make it worse, so I rolled back to latest WHQL certified drivers for windows 7

http://community.bistudio.com/wiki/arma2.cfg

might want to disable/tweak it a bit.

[domokun 515]

don't know what ATOC value is, drivers are all up to date.

Before you start troubling yourself with drivers, try these tweaks

With a bit of luck you should be able to keep your Draw Distance high enough to fly.

[d3lta 10]

Configure to use 4 cpu cores, this helps....

[kotov12345 10]

First of all on first look it seem that your mobo not really fast to have more fps.

Second

in drivers set all AA and AF to application mod.

third in game try set all very high except AA and PP to OFF and play mp warfare on server with other people.

Sometimes I have 25-30 fps with 1280x1024/AA normal all rest very high with faster chipset-mobo (x58 + i7 930) and similar graphic card (HD 4870X2).

Forth if you play SP warfare your fps drop due to your pc works as server and client.

you can change using increasing/decreasing amount of ais resistance town etc.

Edited January 5, 2012 by kotov12345

[maninblue 2]

Configure to use 4 cpu cores, this helps....

Isnt this supposed to happen automatically?

[PuFu 4600]

First of all on first look it seem that your mobo not really fast to have more fps.

please stop spreading that sort of BS, will you.

secondly, i always wonder, since you seem to be living in london, how come your english is sooo poor...

Isnt this supposed to happen automatically?

yes, it should

Edited January 5, 2012 by PuFu

[magnus28 1]

Configure to use 4 cpu cores, this helps....

how about 6 cores?

I have a hexcore AMD 1090t

---------- Post added at 10:31 AM ---------- Previous post was at 10:23 AM ----------

First of all on first look it seem that your mobo not really fast to have more fps.

.

This has little if anything to do with FPS.

If I took the same hardware and changed only the motherboard to the best on the market, I probably would only see 1 or 2 FPS different.

CPU/GPU/MEM/Hardrive, thats it for affecting fps.

unified memory architecture, background programs running, state of HD, etc are all products of the big 4.

Changing mobos does not significantly change anything.

The 890FX chipset is Plenty fast. USB 3.0, 16x pci express, etc.

all the bells and whistles.

[kotov12345 10]

since you seem to be living in london, how come you english is sooo poor...

yes, it should

call me 0044 208 180 76 01

I tested many hardware due to part of my job.

---------- Post added at 03:56 PM ---------- Previous post was at 03:52 PM ----------

If I took the same hardware and changed only the motherboard to the best on the market, I probably would only see 1 or 2 FPS different.

CPU/GPU/MEM/Hardrive, thats it for affecting fps.

unified memory architecture, background programs running, state of HD, etc are all products of the big 4.

Changing mobos does not significantly change anything.

The 890FX chipset is Plenty fast. USB 3.0, 16x pci express, etc.

all the bells and whistles.

ok try take intel h67 chipset place cheapest intel CPU for it and see FPS.

Set 10km VD all normal except shadows and pp and take SU34 flay over chernorus tell me what FPS you will get.

[magnus28 1]

I didn't say I could replace my chipset with the CHEAPEST chipset on the market, I said If I were to take my CURRENT set up, and replace it with the best chipset on the market, I wouldn't see any difference.

I have an FX chipset, which means that if a single GPU is used, it will support 16x PCI Express bus speeds.

If I were to get a cheap GX chipset, it might only go 8x/8x. but mine is 16x no matter how many GPUS I use.

You test hardware as a part of your job?

good for you, you might also know that testing FPS is not a finite science.

I could test my settings looking at a fixed point in the game, then test an hour later, and see 5 fps in the difference.

I would like you to post SS of your test setups that you performed to PROVE that changing a chipset will consistently produce better/worse results.

and I mean compare a FX chipset with a comparable FX chipset.

[kotov12345 10]

Your mobo is 785g chipset,not fx and gx etc :)

http://en.wikipedia.org/wiki/Comparison_of_AMD_chipsets

2009 mostly sell in 2010 not very new.I'd say.I really doubt if you can have more fps with current settings.

Try to do what I ask you to do.

Set in drivers AA and AF into app mode.

Set Vd to 10km and set all setting in game to normal and do PP and shadows OFF.

Take in editor SU34 and fly - what minimum and maximum fps you will get ?

Edited January 5, 2012 by kotov12345

[deszythe 1]

First of all on first look it seem that your mobo not really fast to have more fps.

2009 mostly sell in 2010 not very new.I'd say.I really doubt if you can have more fps with current settings.

this is not true !! has nothing to do with FPS !!! (maybe 1% or 2% , if at all) --- where have you gotten such wrong and missleading knowledge ???

as long as PEG 2.0 x16 is supported, and as long as CPU is supported with MoBo this ---> every thing is perfectly fine

Edited January 6, 2012 by deszythe

[kotov12345 10]

check wiki link above

Recently I found nice test based on 570 graphic card and guys uses latest intel and amd mobos and cpu.

I know it is russian test - but easy to find section where games tested.

You can see that same year intel systems around 20% faster than amd with same graphic card and ram.

http://www.ixbt.com/cpu/amd-fx-8150.shtml

To get fast gaming PC first you need to choose chipset and after that mobo.You can have cheapest CPU for it and according your budget graphic card.Ram is not make any difference 4gb or more.

[deszythe 1]

20% is MAYBE possibel, BUT first this is not that much, and second what you link to is very likely the CPU the biggest factor, not the MB-Chipset. you cnanot compare AMD to intel, you must compare AMD to AMD and with same CPU of curse!!! same system, only different motherboard, any other compare is not valid as indictation of chipset <-> game performanc !!!

[domokun 515]

20% is MAYBE possibel, BUT first this is not that much, and second what you link to is very likely the CPU the biggest factor, not the MB-Chipset. you cnanot compare AMD to intel, you must compare AMD to AMD and with same CPU of curse!!! same system, only different motherboard, any other compare is not valid as indictation of chipset <-> game performanc !!!

I quite agree that comparing different chipsets (and therefore CPUs) is like comparing apples and pears. But! ... There is one very simple comparison that always triumphs, €, £, $, etc. How much performance do I get from AMD chipset & CPU for €XXX vs Intel mobo + CPU for €XXX. All too many "experts" forget this very basic point.

[deszythe 1]

it dont matter , for exampel if you buy some cheap-ass H61 Mobo for 40$ or you buy ASUS Maximus IV Rampage Extreme (Z68 board) for 400$. rest component is teh same, e.g. i5 2500, 6870 Radeon, 8 Gig RAM and decent PSU. then game performance (fps wise) will not be varying much, if it vary at all !!!

of course its totally different, when you compare this to, for exampel an LG 775 board with LG 775 CPU. even if it is the same clock, it doesnt say anything in compare to other !!! but motherboard itself has nearly no impact on performance, as long it support teh CPU ,and teh CPU is not bottle neck the GPU. then everything must be good !!!

and in the case here, i dont know if board support CPU. if it dont, it may need BIOS update, and if still not, it may well be that the CPU is downclocked because the board cant handle it. in this case of course your are right saying its the motherboard. but IF CPU is supported, the rest dont matter. then it cannot account for great performance loss and especially not for THAT low FPS. it maybe makes a few percent, but never ever that much !

Edited January 7, 2012 by deszythe

[magnus28 1]

Your mobo is 785g chipset,not fx and gx etc :)

http://en.wikipedia.org/wiki/Comparison_of_AMD_chipsets

2009 mostly sell in 2010 not very new.I'd say.I really doubt if you can have more fps with current settings.

Try to do what I ask you to do.

Set in drivers AA and AF into app mode.

Set Vd to 10km and set all setting in game to normal and do PP and shadows OFF.

Take in editor SU34 and fly - what minimum and maximum fps you will get ?

Please stop spreading misinformation and lies, you obviously don't have a clue what you are talking about.

My chipset is 785g huh?

not 890 fx like I stated?

you sure about that?

http://www.twentyfour.info/buy.php?c=477303047

now please go away

[NeuroFunker 11]

guys, don't listen to kotov. i already had a discussion with him about that, he is repeating it everywhere: "I have ton of experience testing mobos" "Good mobo = good fps" and stuff, that guy is retarded, or a troll, or maybe both. It is quite amusing, someone believes in that crap, from all components, that actually affect FPS, he choosen most unimportant, and is sure, it playes most of the role in gaming performance. *head bashing wall*

Edited January 11, 2012 by NeuroFunker

[deszythe 1]

yes, i imagine he trolling. no person can be so stupid, especially not when working in IT ^^

[kotov12345 10]

lol to be honest ASUS M4A8TD PRO is office pc mobo with onboard graphic designed to be work in office produce spreadsheets and typing word documents :)

You can play on it - but no way you can have decent graphic performance.

[magnus28 1]

wow! Captain Misinformation strikes again!

According to you in your previous post I was running the 785g chipset, now my mobo is for office application????

Yeah.. because they want to make sure to give all their employees an FX chipset for the intense graphics they will be running, oh an all the overclocking tools on the motherboard for their personal experimentation right?

I think you are one of 2 things Kotov,

either

1. incredibly stupid,

2. incredibly sarcastic

I vote for number 1.