• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

GTX 480 SLI idle temps

Flanno

Flanno

Flanno

Soldato
Joined
18 Jun 2005
Posts
3,434
I moved back to Nividia the weekend as I had no end of issues with ATI that just made my system unstable. I'm now running a pair of gtx 480's in sli on a rampage 3 with a Corsair 600T case on a single 3008WFP screen and am concerned about idle temps.

Initially I noticed that my top GPU was idling at 71 degrees and bottom one at 51 degrees. This is the temp is settles down to after a day of gaming, internet use etc.. Ambient temps in the room are not much more then 18C. Now I have removed the top HD cage on the 600T when doing the build, so the front fan does provide more cool air to the bottom GPU so it's a given it will run cooler. So I would expect the top GPU to run maybe 7-10 degrees warmer but not 20. Interestingly neither go over 95 running Heaven by the way. Also, I've made sure they are running in single monitor mode with adaptive power settings. Both GPU's clock down properly as expected.

Anyway, I swapped GPU positions this morning. I expected the bottom GPU which runs at 51 idle to now run 10 degrees hotter in the top position. It does. I'm looking at it now and after a couple of runs of Heaven, it's idling at 58. The other GPU which ran at 71 idle in the top position is now in the bottom position. This is now running at 50 degrees. So there is a 8 degree difference between both GPU's.

What I can't understand is how this translates into a 20 degree difference when the GPU's positions are swapped ?
 
It's probably the difference in leakage between the GPU's.
There have been allot of reports of large variations from chip to chip.

The issue causes lots of conflict in forums, because some folks say they are fire breathing beasts and others say they are cool even when overclocked...
 
I can't really comment, but my cards idle at 30oC and both with in 1oC of each other under load, but mine are WC.

Before this though I had to sell one of cards cus it was 13oC higher underload and buy another one, so basicly one card was full 13oC higher in the same loop.

Have you checked MSI after burner to see if both cards clock down ok and the difference in the VID.
 
Ok, I monitored another heaven run using MSI afterburner.

Top card maxes at 94 and the bottom card (which used be the one idling at 71 in the top slot), maxes at 89. This is within specs.

On cooldown, the top card seems to cool down a little faster, but eventually it evens out and the top card remains 5-7 degrees warmer then the bottom card as expected. I've just finished a heaven run and they are both at 68 and 60 degrees respectively. GPU's are clocking down fine to 51mhz. Edit : Both cards are now at 63 and 55.

I checked the voltages with afterburner and gpu-z. When both GPU's are running idle at 155mhz the VDCC is 0.95v. When they are under load (400mhz or 726mhz) the VDCC on the top card (which used idle at 51 in the bottom slot) rises to 1.075v. The VDCC on the bottom card (which used idle at 71 in the top slot) is 1.025v. Therefore there is a difference under load of 0.05v. I don't think this makes much difference because the one with the lower VDCC is the one that used run at 71 degrees idle in the top slot!!.

I'm baffled. I suppose as long as the cards don't go over 95 degrees, and games don't crash I don't have an issue. Headphones sorts out the noise.

Should I be using Afterburner fan profiles. Currently letting the drivers handle the fans.
 
Flanno said:
Ok, I monitored another heaven run using MSI afterburner.

Top card maxes at 94 and the bottom card (which used be the one idling at 71 in the top slot), maxes at 89. This is within specs.

On cooldown, the top card seems to cool down a little faster, but eventually it evens out and the top card remains 5-7 degrees warmer then the bottom card as expected. I've just finished a heaven run and they are both at 68 and 60 degrees respectively. GPU's are clocking down fine to 51mhz. Edit : Both cards are now at 63 and 55.

I checked the voltages with afterburner and gpu-z. When both GPU's are running idle at 155mhz the VDCC is 0.95v. When they are under load (400mhz or 726mhz) the VDCC on the top card (which used idle at 51 in the bottom slot) rises to 1.075v. The VDCC on the bottom card (which used idle at 71 in the top slot) is 1.025v. Therefore there is a difference under load of 0.05v. I don't think this makes much difference because the one with the lower VDCC is the one that used run at 71 degrees idle in the top slot!!.

I'm baffled. I suppose as long as the cards don't go over 95 degrees, and games don't crash I don't have an issue. Headphones sorts out the noise.

Should I be using Afterburner fan profiles. Currently letting the drivers handle the fans.
Click to expand...

Are you sure you got the cards the right way around?

If so, then maybe the TIM under the heatspeader may have gaps in it. GT200 suffered with this issue.

gt20065nmsam.jpg


http://www.semiaccurate.com/2009/07/23/how-peek-chip-guts-without-killing-them/

The only thing else I can think of is that the thermal paste might not be spread evenly under the heatsink.
 
Absolutely. Double checked.

I don't want to void the warranty by removing the cooler. It's possible the thermal paste is not spread properly but if I have nothing to worry about, I don't think I will look into that just yet.

Just to confrim since I swapped the cards around they are at 58 and 51 idle which seems fair enough. I am only using the nvidia drivers to control fan speeds so maybe the temps could be lower. I have seen people boasting on other forums of their clocks idling in high 30's on air and I don't know how this is possible even. I'm assuming they are running aggressive afterburner fan profiles to achieve this.
 
Fair enough.

Is the behaviour I am seeing normal though ?

Also, if I do something like run GPU-Z, the clocks on both cards will suddenly go up. Just did it now, and my top gpu went to 70 degrees and bottom one to 60. They were at 63 and 55. It seems like anything at all sets them off. I've exited gpu-z and now they are dropping slowly down.

Regarding the vid issue, I've done some searching and it's common to get cards with different vids. People have reported evga gtx 480's with 1.015, 1.025, 1.075 and even 1.1.

I suppose I could return the card with the 1.075 in the hope of getting one with 1.025 or lower. But is there any point ?

Raven - what temps were your getting on air and what are your vids ?

thanks
 
Interesting.

Are you using msi fan profiles or just letting the driver do the work.

Can you tell me what MSI Afteburner reports for your temps when cards are idle ?
 
Does water-cooling or installing a third party cooling solution on my video card void the warranty?

Answer / Solution
Installing third party cooling solutions does not void warranty on our products. Just be sure to keep the original cooling solution as it will have to be on the card if it is ever sent in for RMA. Any physical damage such as burn marks or damaged PCB will void ALL warranties.
Click to expand...

EVGA FAQ for the US and the rest of the world I would assume.

My temps idle are about 40c/80c load with max fan of 70% and 1.075v, EVGA SC.

I suppose I could return the card with the 1.075 in the hope of getting one with 1.025 or lower. But is there any point ?
Click to expand...

No point, that vid applies when the GPU is maxed out, for idle and low power state the volts drop, you can monitor volts with MSI. You can set volts for each GPU with MSI anyway.
 
Last edited:
Thanks.

I can't understand how my idle temps are so pants to be honest.

Does sli increase idle temps ? I read somewhere it does, but I thought that was with more then 1 screen.

Have you got a pic of your fan profile ?
 
Fan profile




fanhr.jpg










Have you checked your cards are running at idle clocks, mine being 51Mhz GPU, also having any flash content like youtube will raise the clocks to lower power state about 400Mhz.
 
Thanks mate.

Yep. They are running at 51mhz. I have this webpage opened now, and email. Cards are running at 59 and 51 at 44%. I think the issue is down to the pants cooling in the 600T. It just isn't that great. That 200mm fan on the front is the only intake. Thinking of reversing my H70 and making it an intake. It is currently in push/pull withg vipers exhausting.

Anyway, I may look into the HAF-X or FT02.

Edit : I've just applied your fan profile. The fans were running at 40% each and gpu's at 59/51. Now the fans are running at 48 and 45%, and GPU's are at 57/50. So a small improvement for not much extra noise. I think I really need to look into better cooling for these cards.
 
Last edited:
^^^
TBH Flanno, I can't see that the temps are a major issue, unless noise is an issue.
Or is the issue more to do with the fact that one GPU is outputting more heat than the other one, just curious is it you who has OCD like tendencies or am I thinking of someone else?
 
You must log in or register to reply here.
Back
Top Bottom