• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

XFX 6870 - "VPU Recovered " Issues

Z

Zarni

Limp Gawd
Joined
Dec 21, 2005
Messages
160
I purchased the card from newegg in the end of June and started getting the issue right away. It replaced a Sapphire 5850 that I sold on ebay. I am now missing the 5850 as I didn't have any issues with it .

I have tried the 6870 in 3 separate machines*,tried drivers 11-6 to 11-2 , 2 of them got clean OS installs in the troubleshooting process and I have tried WinXP and Win7.All bios and driver settings at defaults.

They all exhibit the same issue, everything looks and runs well until all of a sudden the game screen goes black then the VPU recovered message comes up. UT3, Section 8,Shogun Total War Demo & Battlefield Bad Company 2 are the games I have been testing with.

Tonight I borrowed a 3850 from a friend and everything seems to be doing fine. Even Duke Nukem is running smooth.

So it does seem to be the card to me but I am open to Ideas:confused:

After the first reformat on the machine the 5850 was in, I opened a ticket with XFX. I got a response in about an hour, but now I haven't heard anything in 2 days:mad:

With XFX's slow response I am starting to think that I should just send it back to newegg while I still have time...


*Player1: Asus P8P67-Pro, 2500k cpu,Hyper 212+ heatsink,8GB G.Skill ram,GPU (TBD),60GB Agility2 SSD ,600watt ocz mod xstream power supply, Antec DF-30 case

Player2: Gigabyte 785 matx, 940be cpu,Core Contact Freezer heatsink,GPU (TBD) 4GB OCZ ram, thermaltake tr2 600watt power, supply 500gb seagate HD

HTPC: Abit IP35pro,E5200 cpu, nvidia 220 gpu,2gb OCZ ram,2 - 1TB samsung green HD's, antec 430 earthwatts power supply
 
Last edited:
The drivers are crashing, usually as a result of the card overheating, in turn usually a result of overclocking too high. So if this happens regularly at reasonable temps and normal speeds, then the card is defective.

Could you log with GPU-Z while this occurs to see if anything in particular is out of whack? The card's probably bust anyway, but if you have a specific cause it could help speed up the RMA.
 
To rule out possible in-compatibilities between programs, during the format/re-installs I only installed what was necessary for the card and games to work.

When I first got it though Afterburner was already installed and I was using that to monitor temperatures etc.
After the first few times the errors popped up I set the fans to %100 but that didn't help.
The graphs showed temps in the 70's then the drop in the clock rates & voltages when the driver recovered.
I am starting to think it may be a un-cooled component on the PCB having a issue.
 
Sounds like a lemon. Return it. XFX should have you covered under warranty, but make sure you register it now (at least with EVGA you have to register within 30 days).
 
Thanks for your opinions,I haven't had issues like this in a long time and thought maybe i was missing something.

I made sure to register it before I put in for the support ticket because I would rather rma it with xfx than newegg. Unfortunately with only getting one reply to the ticket so far it looks like it may take too long to be practical.

I did send a PM to the [H] xfx support rep to see if something happened to the ticket.
 
My friend has been having similar issues with his Sapphire Radeon 6870, and just reformatted his computer last night to see if it fixes the issue by reinstalling from scratch. Mind you this build is rather new-- about three months old.

He's tried every driver since 11.3 and is hoping 11.6b Hotfix would not give him problems. He told me last night that if these issues come up again, he's going to see if he can RMA it or replace it. If not, he'll just get an Nvidia card.

I suggested that he do a BIOS edit with the clocks and memory clocks to see if it helps by keeping them the same across all 4 power states because it works for me. Unfortunately, he doesn't want to do that since he doesn't want to risk losing the video card. I've done it to a 5770 and now a 6950. I'm sure a BIOS edit would fix these issues because I've seen similar issues here on various posts from as far back as the 5000-series were released.
 
I was getting that problem with my non-o/c'd gigabyte reference 6870 when I moved to my new case. I would get the crash/recovery every 10-15 minutes. I used HWinfo32 and saw that my memory was running 10c higher then everything else on the card at 88c. I replaced the stock cooler with the twin turbo pro, this dropped my temps 20c, have not had a problem since, the memory also runs the same temp as the core now.
 
Got a response from XFX today and they issued an RMA. I'll update with the results.
 
Oh great my same card has been doing this same thing. I thought it was with overheating but I started to crank the fan speed higher. I guess I will get in touch with xfx :(
 
You must log in or register to reply here.
Back
Top