Crazy temperatures and shutdown in triple SLI GTX580
Hello nvidia enthusiasts, I have quite the problem.

My friend runs a triple SLI EVGA GTX580 setup and we've been seeing a crazy problem. His system uses an i7-930, ASUS P6X58D Premium and a 1200W Corsair PSU. All 3 cards and the CPU are cooled via Aquacomputer blocks. At idle, the first card has a temperature difference of about 20C compared to the others, and when we push them (Heaven 2.0, Furmark, etc.) the temperatures spike and the entire system shuts down. Thanks to an external Logitech LCD (Space Pilot Pro) and EVGA Precision, we know the card shuts down at about 99C, which is right in line with the standard temperature threshold. Removing SLI and adjusting the monitor output to another card, we see the exact same behaviour. Additionally, we're very certain that the cards aren't actually reaching 99C, as the water temperature is near room temperature and the cards are only warm to the touch.

We've contacted EVGA and they haven't really responded. I think it's something to do with the temperature sensors, but I have no idea how to pinpoint that. Drivers are current. What else could it be?
Hello nvidia enthusiasts, I have quite the problem.



My friend runs a triple SLI EVGA GTX580 setup and we've been seeing a crazy problem. His system uses an i7-930, ASUS P6X58D Premium and a 1200W Corsair PSU. All 3 cards and the CPU are cooled via Aquacomputer blocks. At idle, the first card has a temperature difference of about 20C compared to the others, and when we push them (Heaven 2.0, Furmark, etc.) the temperatures spike and the entire system shuts down. Thanks to an external Logitech LCD (Space Pilot Pro) and EVGA Precision, we know the card shuts down at about 99C, which is right in line with the standard temperature threshold. Removing SLI and adjusting the monitor output to another card, we see the exact same behaviour. Additionally, we're very certain that the cards aren't actually reaching 99C, as the water temperature is near room temperature and the cards are only warm to the touch.



We've contacted EVGA and they haven't really responded. I think it's something to do with the temperature sensors, but I have no idea how to pinpoint that. Drivers are current. What else could it be?

Endeavour - i7-950 @ 4.0GHz / 2x EVGA GTX470 / Obsidian 700D / AquaComputer + EK Waterblocks / 12GB RAM / 120GB Intel 510 SSD

MacBook Pro - 13.3" / 2.26GHz / 120GB SSD / 8GB RAM

Ubuntu File Server - 6.75TB / MythTV / AFP / SSH / SMB / DHCP

www.wesg.ca

#1
Posted 04/03/2012 03:32 AM   
I would say the waterblock was not correctly installed onto the GPU.

The GPU heat sensor is within the GPU die, you can't get to it.

The temp of the liquid in the loop being cold does not mean that other components aren't blazing hot.
I would say the waterblock was not correctly installed onto the GPU.



The GPU heat sensor is within the GPU die, you can't get to it.



The temp of the liquid in the loop being cold does not mean that other components aren't blazing hot.
[quote name='-{RaptoR}-' date='02 April 2012 - 11:26 PM' timestamp='1333430800' post='1391288']
I would say the waterblock was not correctly installed onto the GPU.

The GPU heat sensor is within the GPU die, you can't get to it.

The temp of the liquid in the loop being cold does not mean that other components aren't blazing hot.
[/quote]

Indeed, he speaks the truth.

I've run 3 580s in 3-Way SLi heavily OCed (holds the current forum record for Vantage score), on air before and never seen temps that high, so something is seriously outta whack. Either the blocks are loose, the thermal paste wasn't done right, or the cards are faulty. I'd get on eVGA's arse & get them to start the RMA process ASAP.
[quote name='-{RaptoR}-' date='02 April 2012 - 11:26 PM' timestamp='1333430800' post='1391288']

I would say the waterblock was not correctly installed onto the GPU.



The GPU heat sensor is within the GPU die, you can't get to it.



The temp of the liquid in the loop being cold does not mean that other components aren't blazing hot.





Indeed, he speaks the truth.



I've run 3 580s in 3-Way SLi heavily OCed (holds the current forum record for Vantage score), on air before and never seen temps that high, so something is seriously outta whack. Either the blocks are loose, the thermal paste wasn't done right, or the cards are faulty. I'd get on eVGA's arse & get them to start the RMA process ASAP.



Image


Image

Help fight Cancer, Alzheimer's and Parkinson's Disease by donating unused CPU and GPU power to Stanford University's Research Folding@Home projects:

Simplest method is to setup the FAH v7 client with this Windows Installation Guide

#3
Posted 04/03/2012 08:51 AM   
Ouch ! Have you verified that the pump is working ?
Ouch ! Have you verified that the pump is working ?

Image

#4
Posted 04/03/2012 09:35 AM   
[quote name='wesg' date='02 April 2012 - 11:32 PM' timestamp='1333423958' post='1391268']
Hello nvidia enthusiasts, I have quite the problem.

My friend runs a triple SLI EVGA GTX580 setup and we've been seeing a crazy problem. His system uses an i7-930, ASUS P6X58D Premium and a 1200W Corsair PSU. All 3 cards and the CPU are cooled via Aquacomputer blocks. At idle, the first card has a temperature difference of about 20C compared to the others, and when we push them (Heaven 2.0, Furmark, etc.) the temperatures spike and the entire system shuts down. Thanks to an external Logitech LCD (Space Pilot Pro) and EVGA Precision, we know the card shuts down at about 99C, which is right in line with the standard temperature threshold. Removing SLI and adjusting the monitor output to another card, we see the exact same behaviour. Additionally, we're very certain that the cards aren't actually reaching 99C, as the water temperature is near room temperature and the cards are only warm to the touch.

We've contacted EVGA and they haven't really responded. I think it's something to do with the temperature sensors, but I have no idea how to pinpoint that. Drivers are current. What else could it be?
[/quote][i]

Absolutely concur it is the water-block contact. Even stagnate fluid would cool beyond that. [/i]

-Hooks
[quote name='wesg' date='02 April 2012 - 11:32 PM' timestamp='1333423958' post='1391268']

Hello nvidia enthusiasts, I have quite the problem.



My friend runs a triple SLI EVGA GTX580 setup and we've been seeing a crazy problem. His system uses an i7-930, ASUS P6X58D Premium and a 1200W Corsair PSU. All 3 cards and the CPU are cooled via Aquacomputer blocks. At idle, the first card has a temperature difference of about 20C compared to the others, and when we push them (Heaven 2.0, Furmark, etc.) the temperatures spike and the entire system shuts down. Thanks to an external Logitech LCD (Space Pilot Pro) and EVGA Precision, we know the card shuts down at about 99C, which is right in line with the standard temperature threshold. Removing SLI and adjusting the monitor output to another card, we see the exact same behaviour. Additionally, we're very certain that the cards aren't actually reaching 99C, as the water temperature is near room temperature and the cards are only warm to the touch.



We've contacted EVGA and they haven't really responded. I think it's something to do with the temperature sensors, but I have no idea how to pinpoint that. Drivers are current. What else could it be?





Absolutely concur it is the water-block contact. Even stagnate fluid would cool beyond that.




-Hooks

QUOTE (The Professor @ Oct 31 2010, 04:59 AM)

*Jeremy Clarkson face*



So we must hand it over to our tame PC tweaker. Some say he sticky tapes a block of uranium to his dinner before eating it and that he sucks moisture out of ducks. All we know is, he's called Hooks.



"Eye of the Storm" Window Mod Tutorial <> "Inside Crysis 2" <> Top Tier Water-Blocks 2011 <> SSD Unlimited Storage Tutorial

#5
Posted 04/03/2012 11:51 AM   
I have to agree with everyone else here... Hopefully there hasn't been any damage done to the cards. And good luck with eVGA... as soon as you took off their spec coollers, didn't you void the warrenty?
I have to agree with everyone else here... Hopefully there hasn't been any damage done to the cards. And good luck with eVGA... as soon as you took off their spec coollers, didn't you void the warrenty?

ASUS Rampage IV Black Edition / i7 4930k / 32GB DDR3 G.Skill Ripjaw Z / OCZ Agility 3 240GB / OCZ Agility 3 120GB / ASUS nVidia GTX680 (3-way SLI) / Powered by SilverStone Strider 1500w / Windows 8.1 Pro x64 | Custom Water Cooling (Built with Koolance) - full system

#6
Posted 04/08/2012 11:06 AM   
Not necessarily. eVGA warranties do cover fitting of waterblocks, but only if correctly fitted without damaging the cards. Shouldn't need to invoke the warranty here, though, just refit the blocks.
Not necessarily. eVGA warranties do cover fitting of waterblocks, but only if correctly fitted without damaging the cards. Shouldn't need to invoke the warranty here, though, just refit the blocks.

Core i5 4670K @ 4.3GHz with 4GHz uncore
Corsair H100 water cooler
8GB Patriot Viper DDR3 @ 2400MHz
Gigabyte Z87X-OC mobo
SLI eVGA GTX670 FTW (1227MHz boost 6800MHz VRAM)
Soundblaster Z with 5.1 speakers
112GB OCZ Vertex 3 SSD
240GB OCZ Vertex 4 SSD
Corsair Carbide 500R case (white)
Akasa PowerXtreme 1200W PSU
HP ZR2740W 2560 x 1440 27" IPS panel
Windows 7 Pro 64bit

#7
Posted 04/09/2012 07:34 PM   
Scroll To Top