Crossfire overheating/shutting computer down

Status
Not open for further replies.

sassal

Distinguished
Aug 24, 2011
24
0
18,510
Hey all,

I recently got a new PC with 2 6950's (dirt 3 edition). The motherboard I have is http://www.coolingconfigurator.com/upload/pictures/ASUS_Maximus_IV_Gene-Z.jpg

Now, my idle temp for first card is 40C and 30C for second card but under heavy load (playing witcher 2) the first card shoots up to 85C and then it shuts down the computer. It did it after around ~45mins of gameplay the first time and then ~5mins second time i tried to play it.

I have a total of 2 fans on each card, 1 fan on the back of the case, 1 on top, 1 front bottom and 1 on the side pointed directly at the cards.

I know the issue is with the cards basically touching eachother but i cant do anything about that with this mobo so i need water cooling for atleast the top card since it gets no air flow.

Could someone suggest to me a cheap way to get water cooling for 1 card (as i am noob to it) or suggest another way I can lower the temps?

btw my case is haf x-942

thanks!
 
Actually I would guess it's your CPU not the GPU's that is overheating.85c is warm but not hot for these cards.Their max temp is around 100c.

What happen's when a GPU starts to overheat is they will start throttling aka downclocking.It will reduces the clocks and voltage to avoid overheating.The thing is that GPU's can never cause a shutdown of the PC because they will just keep downclocking.

What's shutting down your PC is the CPU overheating.The throttling technique with GPU's is the same for CPU's but once they reach a critical level and just can't keep cool enough it will shutdown to avoid overheating.

Download HWMonitor and post some screenshots of your CPU temps.

What are your full system specs?(CPU,RAM,etc.)
 
Oh okay well i never thought of that haha.

This is a picture of my system idling and I ran witcher 2 for around 5mins and the temperatures didnt go up very much, i will post a screenshot of it under more heavy load later as I can't afford my comp to crash atm as i need to use it 😛

Specs:
i7-2600k CPU
12GB Kingston DDR3 RAM
Asus Maximus Gene-Z Motherboard
2 x Radeon HD 6950 in Crossfire
64GB SSD primarily for Windows 7
Cooler Master HAF X 942 Case

EkgEb
 
alright well i took a look inside my case and i saw that i had 2 8 pin connectors connected to the bottom graphics card and 2 6 pin to the other, i changed the 8-pin to 6-pin and that seems to have worked for now.

i played witcher 2 for around 30mins and my CPU temp stayed around 50C whilst top card was at 79C max and lower card was around 65C max.

also the temp went down around ~10C once i changed the cables around, will do more testing tomorrow
 
Because it was 8-pin to the psu then 6-pin to the gfx card and thats the only cables i had at the time and i figured they would work. they did but not properly i guess haha 😛

Nah thanks for your help :) it usually is a heating problem but I found it odd with my case and fan set-up for it to be heat-related (besides the cards being too close)
 
It's very normal for the top card to be that hot in a Crossfire.

Your PSU is Modular? Really Shouldn't make a difference if your using a 8pin to 6pin because the GPU can only ask for a certain amount of power.Although the 8 pin can deliever more the GPU can't ask for more.
 



CPU max temp 127c 😱 How is that fine?
 
Those are Motherboard temps.The highest reported temp on the CPU was 51c.I do admit even for a motherboard chipset that is really hot.Should probbably change the thermal paste or look for a new heatsink.

Although from the Max Fan Speeds I think this might actually be a buggy motherboard.Should try downloaded the latest BIOS to see if that fixes anything.I know that when I updated my BIOS it changed a few things in the HWMonitor.

What the model of your Asus mobo?
 

With my 2600k and Asus motherboard the AUXTIN reading corresponds with the A.I Suites CPU temp reading, so if the OP's is similar...
 


I think it's the mobos and the way that HWMonitor gets it's data from the sensor chips used.
 
Hey guys thanks for all the replies!

After you pointed out that the max temp was 127 i went and tested it quickly again as i had to leave for uni this morning (i live in aus 😛)

Anyway, i loaded up witcher 2 for 10mins and watched the CPUTIN temp readings, on HWMonitor they never went above 60C (both min and average) but sometimes the min would jump suddenly to 127C but then go back to 60C. So im guessing this is an error with the readings?

also, the other readings were doing the same and looking at the max rpm as being 60k, i think HWmonitor is giving very false readings lol

one more thing, i have the back case fan connected to a CHA_FAN socket, aswell as the side case fan and top case fan. i did this because i dont have another molex cable atm. would that be giving the false readings/overheating possibly?
 
Okay well i cant try that for a while as im out for the whole day so yeah.

If the bios update doesnt change anything, what else could be causing the false temperature readings?
 
Status
Not open for further replies.