cancel
Showing results for 
Search instead for 
Did you mean: 

3rd GPU not recognized on cold start

hunuok
Level 7
Hello to everyone on this forum.

I have a strange issue involving my rig (see sig) on cold startup.

On cold boot, display is not shown in surround (ie 3 monitors).

After login, Lucid Virtu crashes, NV Control panel crashes and only two GTX480s (VGA in PCIe slot 4 not recognized) are recognized in Device Manager. GPUs are seated in PCie slots 1, 2 and 4 as per manual.

If I scan for new hardware changes in Device Manager, I get a code 43 error.

Changes to BIOS include OC to 4.6GHZ, XMP mode, AHCI mode with hotswap enabled.

All remaining settings in BIOS are on default or auto.

GPUs are on stock clocks/voltages.

I have tried:

1. Reseating GPUs,
2. Checked power cables,
3. Both 4 pin Molex are connected to mainboard,
4. Driver reinstallation-Chipset, Sound, LAN, NVIDIA drivers-ALL latest,
5. Updated BIOS to 3202,
6. Swapped/checked monitors and cables.
7. Changed/reseated SLI bridge.
8. Reset BIOS to default settings.
9. Uninstalling Virtu and disabling iGPU settings in BIOS.

I have not added any new hardware/software.

The rig worked FINE (including OC) since day 1 of build. Actually, one of the easiest builds I've had since ATI quadfire+Eyefinity fiasco LOL.

Strangely, if I reset the machine after an hour or two, EVERYTHING works as it should. ie, Surround enabled, SLI enabled, all GPUs recognized in Device Manager.

I am at a loss as to what the cause may be. Bad PCIe slot? Bad GPU? Cold start temp issue?

Any input would be appreciated.
Asus Maximus IV Extreme-Z : Intel i7 2600K @ 4.6GHz :
G-Skill RipJaws X DDR3 12800 16Gb : Galaxy GTX480 Tri SLI : OCZ Vertex 3 120Gb x 2 :
Seagate Barracuda XT 2Tb x 4 : WD Caviar Black 1Tb : Silverstone Strider 1500W :
Corsair 800D : Dell U2711 x 3 in NV Surround : Windows 7 Ultimate x64 :

Koolance CPU-370 : VID-NX480 : HX-CU1020V x 4 : RP-401X2 + TNK-501
9,359 Views
14 REPLIES 14

HiVizMan
Level 40
OK that is a little odd. The cold boot thing makes me thing that the issue is actually located in your ram. Don't ask why I even think that I can not tell you except that the cold boot and ram are linked.

Could you run two cycles of memtest and see if your ram passes please. If yes then we can put that to one side and look at other options.

Cheers
To help us help you - please provide as much information about your system and the problem as possible.

@HiVizMan:

Thank you for your prompt reply.

I can confirm 2 passes of memtest without errors.

I don't know if this is relevant, but I noticed that under system in windows, my memory now reads 16Gb (15.9Gb usable).

The 15.9Gb usable is the first time it's shown like this.

Also, my WEI has gone up to 7.8 (up from 7.4).:confused:
Asus Maximus IV Extreme-Z : Intel i7 2600K @ 4.6GHz :
G-Skill RipJaws X DDR3 12800 16Gb : Galaxy GTX480 Tri SLI : OCZ Vertex 3 120Gb x 2 :
Seagate Barracuda XT 2Tb x 4 : WD Caviar Black 1Tb : Silverstone Strider 1500W :
Corsair 800D : Dell U2711 x 3 in NV Surround : Windows 7 Ultimate x64 :

Koolance CPU-370 : VID-NX480 : HX-CU1020V x 4 : RP-401X2 + TNK-501

HiVizMan
Level 40
And your problem is gone as well?
To help us help you - please provide as much information about your system and the problem as possible.

And your problem is gone as well?


No, I don't think so.

As mentioned, Surround, SLI etc works if I reboot after a few hours.

I won't know until I start up tomorrow morning.

But I'm not holding my breath.
Asus Maximus IV Extreme-Z : Intel i7 2600K @ 4.6GHz :
G-Skill RipJaws X DDR3 12800 16Gb : Galaxy GTX480 Tri SLI : OCZ Vertex 3 120Gb x 2 :
Seagate Barracuda XT 2Tb x 4 : WD Caviar Black 1Tb : Silverstone Strider 1500W :
Corsair 800D : Dell U2711 x 3 in NV Surround : Windows 7 Ultimate x64 :

Koolance CPU-370 : VID-NX480 : HX-CU1020V x 4 : RP-401X2 + TNK-501

HiVizMan
Level 40
Hey it was worth a try 😛

How old is your PSU by the way. It is a pretty hefty beast so that should not be an issue.
To help us help you - please provide as much information about your system and the problem as possible.

The PSU was purchased when I ordered the rest of my rig in late June 2011.

Do you have any other ideas?

I don't really want to disassemble my rig because it is a MAJOR task to drain, disassemble, reassemble components until I can figure out exactly which component is at fault.

BTW, ALL my SATA ports are in use. 2 SSDs, 5 HDDs and 1 ODD.

Also, all my rear USB ports are full. 1 keyboard, 1 mouse, 1 USB HDD docking station, 1 USB ODD (external), 1 Creative Soundblaster 5.1 Soundcard (external), 1 Webcam and 1 Logitech Racing wheel.

I do have to reset pull power chord, then replug) the ports occasionally as some ports fail to work.

Can this have an effect on things?
Asus Maximus IV Extreme-Z : Intel i7 2600K @ 4.6GHz :
G-Skill RipJaws X DDR3 12800 16Gb : Galaxy GTX480 Tri SLI : OCZ Vertex 3 120Gb x 2 :
Seagate Barracuda XT 2Tb x 4 : WD Caviar Black 1Tb : Silverstone Strider 1500W :
Corsair 800D : Dell U2711 x 3 in NV Surround : Windows 7 Ultimate x64 :

Koolance CPU-370 : VID-NX480 : HX-CU1020V x 4 : RP-401X2 + TNK-501

HiVizMan
Level 40
I think what you need to do is put a little more juice into the PCIe lanes. Those 480 cards were pretty hungry I recall.

Have you connected the additional power connector for the PCIe lanes, and increased the amount of voltage to the lanes in BIOS?
To help us help you - please provide as much information about your system and the problem as possible.

Have you connected the additional power connector for the PCIe lanes, and increased the amount of voltage to the lanes in BIOS?


Yes to power connector. No to increase voltage.

If I want to increase voltage, which setting in BIOS and how much?
Asus Maximus IV Extreme-Z : Intel i7 2600K @ 4.6GHz :
G-Skill RipJaws X DDR3 12800 16Gb : Galaxy GTX480 Tri SLI : OCZ Vertex 3 120Gb x 2 :
Seagate Barracuda XT 2Tb x 4 : WD Caviar Black 1Tb : Silverstone Strider 1500W :
Corsair 800D : Dell U2711 x 3 in NV Surround : Windows 7 Ultimate x64 :

Koolance CPU-370 : VID-NX480 : HX-CU1020V x 4 : RP-401X2 + TNK-501

xeromist
Moderator
I know draining the loop to troubleshoot tri-sli is a pain (trust me, I did it no less than 4 times during a LAN party once) but you really need to test that bottom card in a different slot to determine if the card or the slot is at fault.

On a related note, after that LAN party incident I now have Koolance quick disconnects in my rig and they are AWESOME! You have to have enough pumping power to keep flow rates up but being able disassemble my loop without draining it is the best thing since sliced bread. I can even remove the GPU and reconnect the loop without it if I ever needed to install an air cooled card.

The only other thing I can't tell if you've tried is swapping the PCI-E power connectors with the 2nd card. That would tell you if it's a problem with the PSU.
A bus station is where a bus stops. A train station is where a train stops. On my desk, I have a work station…