ArmA3 may have fried my graphics card...

[freddyk83 10]

Hello All,

As the title suggests, I fear that ArmA 3 may have fried my graphics card.

I wouldn't be too concerned about it because 'stuff happens' but oddly enough, this is the second card that an ArmA game has taken out.

This recent card was a

1GB Radeon HD 5670

Not a great card I know but I run a budget rig.

However, even with that card I could run ArmA 3 at a playable rate (admittedly, I run it on default settings so probably not high, certainly not highest!).

The last time I played (Saturday) I was happily playing, for about an hour I guess and then it froze.

I was able to return it to desktop where windows reported driver had failed.

Later that day I ran it again, again for about 1-2 hours, and then it froze again, but then screen went black.

At that point, I had to hard reboot it but it simply wouldn't boot up, even to BIOS.

Monitor reported 'No signal'.

So I remove the radeon, plug the monitor into the onboard and it boots up fine so that pretty much says to me the card has gone to 'silicone heaven'.

As I run a budget rig, this weekend I have sourced a replacement card - a 2GB Radeon HD 6670, the model up i'm guessing (tbh, I don't know how these numbers work).

it's a GB more than my previous card so that should count for something I hope.

My real point for posting this is not to rant at how ArmA keeps killing my cards because I know its not the games fault in fairness.

They ARE graphical demanding games and my card was about 2 years old so it had a good innings.

However, I AM a bit concerned that I don't go and fry THIS one one week into having it! lol

While I love ArmA and am longing to resume playing it, it would be foolish of me to blindly attempt to play it with a card that wouldn't handle it (AT ALL, forget max settings, just running it).

Better to be safe first and ask!

So my question is, is this a real problem that is likely to occur [having a card burnout in a week if I run ArmA3 on what I presume is Low-Mid] or can I relax a bit and enjoy some quality ArmA time again?

And...

Are there any measures I can take to help prevent potential repeat burnouts?

I'm guessing [hoping] it realistically takes a while for a card to burn out and also, can most likely be prevented.

Full System Specs

------------------

Intel Core 2 Duo 3.06ghz

2.1 GB Ram

Windows 7 Ultimate 32-bit

ATI Radeon HD 6670 2GB DDR3

[jumpinghubert 49]

Are there any measures I can take to help prevent potential repeat burnouts?

yes.........the temps :p

[RuecanOnRails 10]

It could perhaps be your power supply.

I don't see Arma as being the culprit. Rather just coincidental that it died while running arma, could have happened with any demanding game.

Searching I found this http://www.evga.com/forums/tm.aspx?m=874078&mpage=1

Not directly the same issue, but it is about someone who had a couple cards fail on them within months. Most people there also speculate the issue was from the power supply.

Do you have access to another computer you could try the "dead" card in to make sure it's truly is dead?

[fabio_chavez 103]

apparently, since the beta, i got increased fps but also increased core temps, at first i thought my aircooling reached its limits because when the beta came, it also was really hot in my town.

But now the local temps dropped but my CPU temps are still above normal , i just meassured 97°c!!! even for my oced xeon that is WAY beyond anything i ever experienced before and far above tollerable level.

I recoded a vid with 12k viewdistance and 200% sampling, though it shouldnt result in a china syndrome... -_-

not 100% sure if the BETA is the cause but i recommend to awareness, especially if you have to squeez every last mhz out of the cpu to be able to play the game, like me.

[mr burns 132]

Once killed a 8800 GT whilst making screenshots for BWmod at fullest possible quality (true story).

And in that case i already kinda knew it was dying (i.e. forced) because it started showing full screen normal map textures during FO3 gameplay before.

Just wanted to point out: ArmA doesn´t kill GPU´s, users &/or HW faults do :D

[freddyk83 10]

Do you have access to another computer you could try the "dead" card in to make sure it's truly is dead?

Unfortunately I don't.

However, one thing I DID do was when I first removed the card and managed to boot it up using onboard, I then shut it all back off and re-inserted the card (my theory being sometimes taking things out and then slotting them back in again sometimes tricks them into working) and the first time I tried that, it worked!

I actually got back into Windows using the Radeon...however, about 5 mins later, the screen went all 'patterny' and the monitor shut off again.

Subsequently went back to only booting up with the card removed and no further 're-insert' attempts have worked.

[RuecanOnRails 10]

When it gets "patterny" that's video memory corruption. Likely a dead card.

Do you have a surge protector or is your computer plugged directly into the wall?

[freddyk83 10]

When it gets "patterny" that's video memory corruption. Likely a dead card.

Do you have a surge protector or is your computer plugged directly into the wall?

It's plugged into a multi-plug adaptor. I don't know if the adaptor is surge protected though.

Think it's just a standard multi-socket.

[13islucky 10]

I actually was getting the same issues on ArmA 2 until I put -nobenchmark in my launch parameters. It might work but I can't be completely sure.

[cifordayzserver 119]

Get non conductive heat paste and re do the thermal paste, and get heat dissipation pads for the memory, you will likely be fine. Or you'll be out 10-20 bucks and KNOW the card is fried.

I've brought 2 or 3 cards "back from the dead" by doing that... budget rigs often are running at 100pct for much of the time, which leads to constant high temps which leads to breakdown of paste, which accelerates the problems and can mimic outright failure, but it's usually just overheating

[Tonci87 163]

1. Check your Power supply

2. Never try to save money on the Power supply, it is one of the most important pieces in your PC. A bad PSU might fry other parts and then it gets really expensive.

3. Give the card to a repair shop and let them test it.

[jumpinghubert 49]

sorry I have to do again captain obvious: the first thing is to check the temps in your case/of your gpu to suspend an overheating issue.

[NeuroFunker 11]

so, you don't even know how to measure themps of your graphics card, but you wanna claim, arma 3 has burned it? OK...

[jumpinghubert 49]

so, you don't even know how to measure themps of your graphics card, but you wanna claim, arma 3 has burned it? OK...

exactly this was my first thought lol

[EDcase 87]

Related to temps; Its VERY important to make sure all fans are dust free. Its critical to have good airflow through the computer.

Also as mentioned before, having a good power supply is more important than many people realize.

ARMA may have pushed the card to its death a bit faster due to load/heat but software cannot kill hardward. (Unlike the old days [bBC micro cassette drive] ;))

Graphics cards and processors have auto shut down in case of serious overheats so excessive heat should not kill a component.

BUT constant use at high temps near operation limit WILL reduce the component life.

Edited June 25, 2013 by EDcase

[chrisb 196]

Run msi or something similar in the background, have your temps displayed (I have mine on the small screen on my G19).

That way you don’t fry anything, 97c is a little too much, although they can hit 100c I’ve been told by the ‘fire brigade’ ;):p.

I never let my card get above 75c, usually hovers around 65-70c on heavier missions.

Also you would have been better going upto a 68** or 58**, it’s the second number with these cards, there won’t be that much difference in performance with your upgrade, but it may handle it slightly better, although a 2gb or above is favourable for this game.

:)