Asus GTX465 SLI cards with different voltages

[BlindObject]

So I have two GTX465s, and normally they're suppose to be idling at .88V and they peak at about .97. However, one card is always at around .96 and the other is at .88. The odd card is also running a little warmer. I tried MSI afterburner. It helped it idle at .88, however, after the odd card reaches the peak again, it won't go back down to idle core voltage and stays at .96v.

Any suggestions?


Also, soon I'll be upgrading to a i7 system soon, but I'm unsure if I should keep my SLI set up, or move onto a single GTX570. After some looking around it appears the 570 has the same horse power as both 465s, if not a little less (without OC). Any opinions?

Current system:

Q9550 @ 3.61ghz with 1.39V
DDR6400 @ 850mhz (4-4-4-12)
Corsair 850W
Asus GTX465 SLI @ 730/1603/1460

 

[BlindObject]

Really? Nobody can help me out??

 

[red1776]

The utilities that read GPU voltages are notoriously inaccurate. Also the lead card may have a small load on it while doing menial things (surfing etc) and may be a little warmer because it is not in a throttled down state as much as the linked secondary card. but that aside, It is not at all strange to have one card (even identical models) run at different temps. reasons can range from having them in SLI (stacked) to different fan speeds and sensors, to how well (or not) the thermal Compound was applied at the factory. If you are not experiencing any problems (crashes,overheating,BSOD's, artifacts, etc, I would not worry about it at all.

 

[dividebyzero]

There are no set GPU voltages for Fermi based cards. I'd be more than a little surprised if you had two cards that operated at the same voltages (Idle/2D/3D) even if they were same batch- or even if the serials followed one after the other.
Unless one (or both) cards are suddenly requiring more voltage to maintain the same clocks- and producing more heat as a byproduct with the fans ramping faster then I wouldn't see it as a problem...although that kind of depends what you mean by "a little warmer".
If you're saying that when the card ramps to 3D clocks and subsequently stays at the same voltage once the clocks drop back to idle/2D clocks then that would imply that either the cards voltage is always at 0.96v (which doesn't from your wording sound like the case), or the card on system start up is showing it's correct idle/2D clocks but gets "stuck" once a 3D app is launched.
Assuming the latter is the case, have you verified the voltages with another utility such as GPU-Z ? -i.e. run 2D > 3D > 2D and monitor the voltage changes (don't initialize Afterburner-use stock card setting/profile)
If you have, then it's safe to rule out an MSI Afterburner bug, and possibly the problem is the cards VRM circuit - a guess on my part as I haven't encountered the same symptoms in my card testing.

 

[BlindObject]

Thanks for the info, it actually helped!

 

[red1776]

what was the issue Blind?

 

[Lord Horns]

I have the same issue with my Galaxy GTX 470's. GPU 0 shows a voltage of .962 when the 3D clocks are going and GPU 1 shows .950, same clock freqs. GPU 0 idles around 50C while GPU 1 idles at around 35-40C. FurMark gets my first card to <108C, card 2 never tops 69C. Then again, my cards are sandwiched, so I have to spot cool the first card for the time being to get some airflow over the heatsink.