Advice needed: 7970 crazy temp within 2 seconds

[27]

n00b
Joined
Apr 16, 2008
Messages
59
My gut on this is that somehow the motherboard is causing this card to misreport the temp it is at, but then again I have no idea why the heck the MB would/could do that.

Anyways, heres the problem. I have a 7970 that I dropped into my sons box which goes from idle of ~46C(normal) to 95C+ in about 2 seconds when a game is loaded (so not normal behavior). Fan on the card also jumps to 100% and stays there. NOTHING on his box is OC'd. MB is a D975XBX2 (badaxe 2).

The same card does not behave like this in my system. I have reinstalled Windows on his system and the problem persists. Occurs both in Win 7 and Win 8.

Any ideas as to what might be causing the issue? Any ideas if it is a different issue altogether?

Thank you,

[27]
 
Place your hand on the PCB and determine if this rise in temperature represented on-screen exists. If your card does indeed rise in temperature, then I would suggest remounting the cooler (and perhaps re applying thermal paste). Your motherboard should have nothing to do with the sensor on your card AFAIK.
 
There is no rise in temp, the exhaust coming out of the card is cool-ish (cooler actually than what you might expect) It would probably rise to a normal exhaust temp if it continued under load though. I am truly stumped on this one.
 
I'd take it apart and see what's going on with the paste.

Edit: And stop using it til you figure it out. Check with another card if you have a spare. If you can't solve it, RMA that sucker.
 
the card works fine in another system though, which (to me) points to some issue either with the OS or mobo
 
could be a bad power supply, do you have a spare to swap out?
 
Did you check with afterburner or one of its sister programs just to double check the readings are correct?

ASUS GPU Tweak
EVGA Precision X
MSI Afterburner
Gainward EXPERTool
Galaxy / KFA2 XtremeTuner HD
Galaxy XtremeTuner Plus
Gigabyte OC Guru II
Sapphire TriXX OC Utility
ZOTAC FireStorm
Palit ThunderMaster

I would be inclined to choose the one that matches the brand card you use...but i think they all written by the same guy...you did say the fan speed jumps to 100% so it kinda seems like you have one of these installed...but maybe amd cards are different like that
 
I would check for BIOS updates on the MB as well as a different monitoring program.

Mac,

A bad powersupply would not be the cause here...
 
Seen it happen, more than once

Bad voltage on rails causing current over-draw.
 
Verify that the heatsink is properly contacting the GPU die.

If anything, it'll at least get you better paste than oem.
 
OP: are the 7970s, GHz editions? If not, are they stock?

This started happening to me a few weeks ago and that's because I flashed a GHz edition bios onto my normal 7970. I will flash it back today as soon as I get home from work and let you know the details.
 
I had something similar happen to a gtx 260 before,it turned out to be bad voltage regulators on the card,it caused the GPU to climb to over 100C and start throttling even at 100% fan speed.
 
I had something similar happen to a gtx 260 before,it turned out to be bad voltage regulators on the card,it caused the GPU to climb to over 100C and start throttling even at 100% fan speed.

That must've been terrifying to you haha.

I've had that happen to my mate's pc, OP, you need to check your 12v rails.
 
Have you tried a different PCI-e slot on the mobo?
Also I'd probably check to see if there is an updated bios for that mother board.
PSU could be an issue as well, if you can swap it out with another for a test, it'd be worth a shot.

The temp sensor should be on the video card itself. And if it works correctly in your machine, that would make me think the sensor is ok, and the problem would be with the monitoring software or a problem with how the mobo is interpreting or transmitting the data to the OS. And of course the PSU may be shooting some freaky voltages to the card causing it act freaky.
 
Back
Top