New 1080ti issues

[GizF1]

Hi all

Since upgrading to a 1080ti (Gigabyte Aorus) and a additional HDD for storage I have been having issues with my PC. It keeps either crashing to desktop, freezing and unable to restart the graphics driver or straight reboot. I also see a fair few DWM errors in task manager.
This is mainly whilst playing Dragon Age Inquisition, Forza Horizon 4 and Plants vs Zombies GW2 but it has happened just on the desktop a couple of times.
I previously ran a Sapphire RX480 Nitro + OC at 1300mhz with no issues.

Spec
i54670K at 4.4ghz
Asrock Z87m extreme 4
16gb (4x8gb) DDR3 at 3400mhz
1 x sata SSD, 2 x sata HDD
Cooler Master G750m PSU (Over 4 years old now)

The only thing that seems to help a little is to reduce the power limit on the card (its down to 50% at the moment) so I am pretty sure its the PSU. When games do play they run very well. I just wanted to double check before I order a new one that I might no be missing something else. So far I have......

Reinstalled Windows (in case it was a switching from AMD issue)
Put the CPU to stock
Updated BIOS
Connected an 8 pin PCI E connector from seperate PSU lines to the card
Downclocked the 1080ti
Reduced settings
Reinstalled and verified game installs

If anyone else could confirm my thinking that its probably the PSU or suggest anything else to try and chuck some brand reccomendations at me that would be great.

Giz

 

[Panos]

Hi all

Since upgrading to a 1080ti (Gigabyte Aorus) and a additional HDD for storage I have been having issues with my PC. It keeps either crashing to desktop, freezing and unable to restart the graphics driver or straight reboot. I also see a fair few DWM errors in task manager.
This is mainly whilst playing Dragon Age Inquisition, Forza Horizon 4 and Plants vs Zombies GW2 but it has happened just on the desktop a couple of times.
I previously ran a Sapphire RX480 Nitro + OC at 1300mhz with no issues.

Spec
i54670K at 4.4ghz
Asrock Z87m extreme 4
16gb (4x8gb) DDR3 at 3400mhz
1 x sata SSD, 2 x sata HDD
Cooler Master G750m PSU (Over 4 years old now)

The only thing that seems to help a little is to reduce the power limit on the card (its down to 50% at the moment) so I am pretty sure its the PSU. When games do play they run very well. I just wanted to double check before I order a new one that I might no be missing something else. So far I have......

Reinstalled Windows (in case it was a switching from AMD issue)
Put the CPU to stock
Connected an 8 pin PCI E connector from seperate PSU lines to the card
Downclocked the 1080ti
Reduced settings
Reinstalled and verified game installs

If anyone else could confirm my thinking that its probably the PSU or suggest anything else to try and chuck some brand reccomendations at me that would be great.

Giz

Did you use DDU to wipe out the AMD drivers first?

 

[GizF1]

Yeah - and then I did a total windows reinstall anyway

Giz

 

[Panos]

Yeah - and then I did a total windows reinstall anyway

Giz

On the 4 extensions on the back of the PSU. 2 going to the GPU without any other devices attached yes?

 

[GizF1]

Yes - what else could you attach to them anyway?

 

[Woodsta888]

HAs a fairly powerful single 12v rail, but as those VGA power leads arent modular, I'd try one of the other two VGA power leads. Only in case that the two you are currenty using, are split in the wiring and one of the other two spares are indeed more seperated to the rail than the ones you are currently using. It's a bit of a hope but worth a try. Seeing a lot of PSU's recently on top tier cards giving trouble.

Use the Windows event viewer at the time of crash to see if that indicates to something.

 

[Resistance]

A bios update would not hurt

 

[GizF1]

Event viewer doesn't really say much just the appcrash or display driver failing to restart and if it reboots it tells me that the last shut down was unexpected. Sometimes there are DWM errors generally when using it for internet or non game stuff. No BSOD codes or anyting

I have done BIOS update - which also gave me spectre and meltdown fixes I didn't really want as well but never mind hey!

Giz

 

[GizF1]

Just to update incase any one Google's their way into this that the problem was the default fan profile on the Aoros is too conservative and the card was getting too hot. Bit of tweaking of the fan curve in afterburner and all is good with the world

Giz

 

[Skynet5]

Just to update incase any one Google's their way into this that the problem was the default fan profile on the Aoros is too conservative and the card was getting too hot. Bit of tweaking of the fan curve in afterburner and all is good with the world

Giz

I'm glad that works, but what are your temps for that to be happening?

Something still looks wrong.

 

[ljt]

it shouldn't be getting that hot it crashes in games on the default fan curve. Is the airflow restricted or something?.

 

[WantoN]

Thought about repasting it? Might be a bad seat or poor TM application.

 

[GizF1]

It was getting up to about 80 degrees I didn't think anything of it because I've come from an rx480 which is quite hot and the temp limit was defaulted to 84 so I thought it was normal.

When I get a sec I'll reset afterburner and see what fan speed it runs at on default.

I'm guessing that the quite high oc the Aoros comes with means it was trying to stay a little bit too fast for the temps because gpu boost was saying 'hey I've got power and voltage to spare here' .

Air was kind of restricted because I was using an matx aerocool dead silence case. I've since changed to a full atx case.

Might look at re TIM in future but happy with noise and temps now

Giz

 

[Radox-0]

the card should still work perfectly fine at 80 degrees and not crash. You will see a drop in the boost clock ( though this starts much lower) then once you hit the target temp which is usually around 83 or so degrees the clocks taper off significantly. Still not normal for it to just crash instead just a heavy drop on clock speeds to stay cool. heck before moving to water, my 1080tis in a sandwhich sat at 83 degrees often