Very high GPU temperatures (M18x R2)

So, my M18x seems to be running unusually hot. Tested it earlier with a 15 minute run of Furmark, and the primary 7970M was peaking at 97C, secondary was hitting 92C. This is after thoroughly dusting the heatsinks yesterday with a can of compressed air.

Are those temperatures normal for sustained 100% load, or are they much hotter than they should be? It doesn't get that high while gaming - it peaks at around 90C primary, 85C secondary while running Sleeping Dogs, which is a very demanding game on the highest settings - but even that seems hotter than it should be.

Is it possible that I've damaged the fans and they're not running as fast as it should be? Or do I potentially need to remove the heatsinks entirely for a thorough clean, and/or change the thermal paste on the GPUs?

Thanks guys.

 

Holy cow... I'm not that familiar with the 7970m cards but that does seem really high to me. Are both of the fans kicking on under load, can you hear them? Did you shoot air through them with the underside cover on or off? I'm wondering if you didn't possibly compact all of the dirt/dust into the heatsinks rather than blow it out entirely. Also, what's the approximate ambient temp where you're running?

 

Thats definitely too high. They shouldn't be close to 80 under the harshest of conditions. I doubt its due to not repasting.

Have you ever cleaned the fans and heatsink before ?

 

Yeah, there's plenty of airflow out of the back too...but the fans potentially aren't as loud as I'd expect under maximum load. Are the GPU fans any quieter than the CPU fan under 100% load?

Dusted with the back cover off. Plenty of dust came out.

Ambient room temperatures are pretty 'normal'...around 24-25C if I had to guess?

I've dusted them fairly regularly, ever since I've got it. I haven't done any 'deeper cleaning' than simply dusting it with a can of compressed gas. These temperatures are definitely higher than I recall them being previously. I think I'll try dusting it again to see if there's any more dust lying around.

 

If you haven't already, I would unscrew the fans and lift them out of their pockets (leaving the wires connected if you wish) and have a look. If you don't remove the fans, anything blown out of the heat sinks with compressed air other than fine dust (i.e. lint and hair) will end up getting blow back into the heat sinks and laying across the surface of the heat sink radiators. This will restrict air flow and that will decrease cooling efficiency. It might not completely block air flow, but any impediment is undesirable and will increase temps.

If the heat sinks are nice and clean, then you're probably going to have to repaste.

 

Well, it's much better already (although not ideal) after another good dusting. After another 15 minute Furmark run, peaked at 86C on the primary GPU and 82C on the secondary. Perhaps I hadn't dusted properly the first time? Not sure, but whatever the reason was, I'm happy with it for now (unless it should still be significantly lower than 86C?) If I feel temperatures remain too high, I'll have a look under the fans.

I was planning on repasting at some point, but I just don't feel confident enough to disassemble my laptop to the extent required to do so. (Perhaps if I ever get around to repasting, I might replace the CPU heatsink for the triple heatpipe model...)

On an unrelated note, I've now got an issue where the secondary GPU seems to be stuck at high load. HWiNFO currently shows it as sitting at 3D clocks (850/1200) and supposedly 99% load, idling at 68C...it's not just a reading error either, I can feel the airflow coming from the secondary GPU fan. Restarting doesn't seem to fix it, only thing that seems to is disabling then enabling Crossfire...which runs the risk of the drivers nuking themselves, as they seem to love to do whenever I try to disable then enable Crossfire. Why didn't I just spend the extra for 680Ms...

 

If you are running one of the newest Catalyst drivers, I think that was a known issue. I have had trouble with clocks getting stuck on AMD video cards in both 2D mode and 3D mode randomly since 6970M CF first came out. I normally didn't worry too much about getting stuck in 3D mode, but getting stuck in 2D mode is definitely a problem.

 

I think you might need to repaste. I am running the latest Catalyst drivers and even with an insane OC with Ambient temps at 30C of Japanese humidity with AC off they only reach 76C.

 

I found this video on re-pasting a GPU, it might help you a lot.

 

My 7970M GPUs reach a maximum of 75°C and 71°C in FurMark for longer periods running Catalyst 13.8 Beta 2. The fans spin up but aren't really all that loud.

 

You need to do re-paste. My card used to reach 90C while playing Battlefield 3. After done the re-pasted yesterday, they never go over 72C

 

is DELL stock TIM THAT bad ? geez...

 

No, stock Dell TIM is excellent and so are the stock thermal pads. Assuming it is assembled correctly at the factory, better than any aftermarket paste or thermal pads. It does an effective job of cooling and it is more durable. But, nothing lasts forever.

FYI - dr_gin_rain's system had been upgraded to 680M SLI with GPUs that were purchased from a member of this forum some time ago and unless he purchased brand new heat sinks from Dell it did not have stock Dell TIM.

 

I don't think I have seen my GPU even with a wild overclock go over 80c...

Normally sits at between 70-75c while gaming and 50c at idle

 

I'm surprised it can be better then something like ic diamond. Anyone know what they use? Brand.. Or material ect

Sent from my SPH-L900 using Tapatalk 2

 

I've tried to find out. I think it is a proprietary product and maybe a trade secret. From what I understand it is not available for purchase and only comes on a new heat sink. The TIM on the die surface is a phase-change pad.

 

I'm curious on the subject too. I know this is an old thread but it's the closest I could find to what I'm looking for. I have a gtx880m sli setup on my 18x r2 and while benchmarking, I got up to 93 Celsius. I know they had issues with heat and throttling so I was not sure if this was normal. I'm using the 144.75 mood Mr Fox provided in another thread. It seems to have disabled throttling but because of the temperatures I have to keep a closer eye to make sure I never reach over 95C. Any feedback would be great thanks

 

I recommend you to do a re-paste as well. Those temps are way too high!

 

OP should consider the removal of the fans in order to properly clean them and the heatsink fins. As Fox stated, just blowing will eventually clog the heatsink with debris that cannot be blowed out. Also a repaste would be nice.

 

Clean the vents... The dust build up causes a lot of temperature problems.. Also repaste your CPU+GPU with some good thermal paste like IC Diamond..

 

Thanks for all the feedback guys! Actually it was a software problem. I ended up using msi afterburner to control the temperatures and it seems to work great! Recommended program for anyone wanting to customize their graphics cards to the extreme! Seriously lol