6990 powers down in any game

sm00nie

n00b
Joined
Jul 8, 2010
Messages
7
Hey all,

I have a 6990 and I purchased it two years ago. I haven't had any issues with gameplay and even though the card ran particularly hot when playing any graphic intensive game that pushed the card to 90-100% load, it was doing well other than the crazy loud fan which I could deal with.

Two days ago however, while replaying Skyrim the card just powered down a few seconds to a minute into the game. Both of my monitors showed that they had gone into power saving mode and that there was no signal (when toggled off and on). I then rebooted, via case switch, and I heard some beeps -- 1 long and 3 short which looking at my motherboards manual indicates that something is up with the 6990. I took the card out, removed any dust/fur and sprayed some air to remove any other dust that I didn't get. I wiped the contact point, and pci-e slot and reinserted the power connectors and plugged it in. Once I loaded up skyrim again, the same issue occured. I then tried the other 2 games I play - simcity and battlefield 3 and the card powered down again. I switched to another pci-e slot and the same issue occurred. I put the fans to 70-100% and the card would still power down minutes into skyrim. I then hooked up an old 4870 and there were no problems with that card. I've tried all the drivers from 12.6 to 13.3beta.

I'm thinking maybe the thermal paste has crusted and needs to be reapplied, but before I do that I thought I'd ask if anyone experienced this issue and have any ideas on what else I could try. I've opened a ticket with Sapphire to see if I can get this card RMA'd, but I'm not sure if it's past the 2 year warranty period just yet (I'm sure it's sometime this month or early next month) and I don't want to have it voided for replacing the thermal paste.
 
You need to monitor the GPU temps up until it shuts down. If it hits over 90c, you've found your problem.
 
Heat or power. Are you overclocking it? How old is your PSU?
It's running at 880/1250 on the second bios. I've also tried the first, both at default values. My PSU is 2 years old, it's a Silverstone strider 1000w (80+ silver).

You need to monitor the GPU temps up until it shuts down. If it hits over 90c, you've found your problem.
The funny part is that this card gets super hot with 99% of both cores running with the automatic fan profile. I've been able to offset it by setting up my own fan profile in afterburner, but even at reaching 60-70 degrees, it'll still power down.

After monitoring usage on both cores, GPU2 never exceeded 72-73% usage in furmark @ 2560x1600 burn in and bench. This was kind of troubling as both cores would always hit 99% usage when I was testing it before.. something is definitely funky. So, I loaded up RadeonPro and disabled crossfire and launched Skyrim and Simcity, both of which GPU1 hit 99% usage and I was able to play each for an hour and satisfied at that point. I enabled crossfire in RadeonPro and gave Skyrim another go and again the card powers down, although the GPU2 usage was at 99%. It seems when it's GPU2 is sustained at 99% usage in game it powers down, even though the temps are hovering around 60-70 degrees with my fan profie ramped up.

I don't think it's the PSU as I'm able to run furmark for a prolonged period, although GPU2 only hits 72-73% usage. I have an ocz 600w psu that I'll use just for the vid card tomorrow to test just in case though.
 
It's running at 880/1250 on the second bios. I've also tried the first, both at default values. My PSU is 2 years old, it's a Silverstone strider 1000w (80+ silver).

The funny part is that this card gets super hot with 99% of both cores running with the automatic fan profile. I've been able to offset it by setting up my own fan profile in afterburner, but even at reaching 60-70 degrees, it'll still power down.

After monitoring usage on both cores, GPU2 never exceeded 72-73% usage in furmark @ 2560x1600 burn in and bench. This was kind of troubling as both cores would always hit 99% usage when I was testing it before.. something is definitely funky. So, I loaded up RadeonPro and disabled crossfire and launched Skyrim and Simcity, both of which GPU1 hit 99% usage and I was able to play each for an hour and satisfied at that point. I enabled crossfire in RadeonPro and gave Skyrim another go and again the card powers down, although the GPU2 usage was at 99%. It seems when it's GPU2 is sustained at 99% usage in game it powers down, even though the temps are hovering around 60-70 degrees with my fan profie ramped up.

I don't think it's the PSU as I'm able to run furmark for a prolonged period, although GPU2 only hits 72-73% usage. I have an ocz 600w psu that I'll use just for the vid card tomorrow to test just in case though.

I don't think it has anything to do with the PSU, but if you have the extra one its worth a shot. I personally think the VRMs are overheating/shot on GPU2..Hopefully it's still under warranty..

If it is and they can't repair the card, chances are they will not have any 6990s lying around..If it happens, I would expect NOTHING less then their highest model 7970 Ghz Edition (new card, not a refurb)..You purchased a $800+ GPU after all...
 
I don't think it has anything to do with the PSU, but if you have the extra one its worth a shot. I personally think the VRMs are overheating/shot on GPU2..Hopefully it's still under warranty..

If it is and they can't repair the card, chances are they will not have any 6990s lying around..If it happens, I would expect NOTHING less then their highest model 7970 Ghz Edition (new card, not a refurb)..You purchased a $800+ GPU after all...

I tried my ocz psu and the same issue occurred :/. If the VRM temp (GPU2) goes up anywhere between 58-62 degrees it does look like it'll power down, so that could very well be it.

I read over their warranty information and it looks like I'll need the receipt and I need to take it back to the original shop I purchased it from. Unfortunately, I don't have the receipt and I paid cash, so there's no record on my cc's either :/. Hopefully, they don't ask for that information when replying to the support ticket :p
 
I tried my ocz psu and the same issue occurred :/. If the VRM temp (GPU2) goes up anywhere between 58-62 degrees it does look like it'll power down, so that could very well be it.

I read over their warranty information and it looks like I'll need the receipt and I need to take it back to the original shop I purchased it from. Unfortunately, I don't have the receipt and I paid cash, so there's no record on my cc's either :/. Hopefully, they don't ask for that information when replying to the support ticket :p

If the RMA falls through, you can repaste/re-themal-pad your 6990 cooler to cool the VRMs better. For good thermal pads I recommend Fujipoly, which you can find in different thicknesses at FrozenCPU or eBay.
 
If the RMA falls through, you can repaste/re-themal-pad your 6990 cooler to cool the VRMs better. For good thermal pads I recommend Fujipoly, which you can find in different thicknesses at FrozenCPU or eBay.

Yeah, apparently just from looking at the serial, they told that the warranty had expired and without a receipt I was boned. I looked at my online statements and there it was in May of 2011. It only lists the purchase price, but I didn't buy anything else at the shop for that price.

I called them up and they told me to bring the card down and they *might* be able to pull up the receipt. I'll drop down tomorrow and hopefully they'll be able to pull something up. They might charge me for the look up, but if they find it it'll be worth it.

Thanks for the thermal pad source btw, I've bookmarked it just in case :)
 
58-62c on the VRMs is very very normal for a 6990. My 6990's vrms used to be at 50c when the card was completely idle. It's not uncommon for them to get close to 100c under load.

There's an issue there somewhere though. Good luck getting it replaced.
 
58-62c on the VRMs is very very normal for a 6990. My 6990's vrms used to be at 50c when the card was completely idle. It's not uncommon for them to get close to 100c under load.

There's an issue there somewhere though. Good luck getting it replaced.

Yeah I hear ya, it used to hit up in mid to high 90's, but now once it gets to 58-62c it'll just shut down and it's specific as to what I'm doing too. If I launch furmark it'll run fine, although the VRM temps will be in the low to mid 50's to low 60's, but if I run a game the vid card will power down and I will be forced to reboot, at which point the card is no longer recognized and I need to do a hardboot to get it recognized again.

I placed the 6990 on my workstation and it also powers down with a 850w gold psu (my backup 600w ocz just powering the card showed the same results) as I was thinking that my mobo may be the issue.

I just hope when it's being tested at the Sapphire RMA center, that they're able to identify the issue or find it at least, so that I don't get send back the same card and with them saying they couldn't find a problem. It's been tested on two other pc's now showing the same result while playing Skyrim (simcity on my workstation as well).
 
Can't be your motherboard or the power supply if the same thing is happening on two different computers. If you get the same card back then you'll know that Sapphire doesn't care.
 
Back
Top