DXGI_ERROR_DEVICE_HUNG error: Is there a solution?

[cj-ibbotson 77]

3 hours ago, SteveW said:

Yes, won't be P3D per-se, will be down to an error in the way a function is called, perhaps with a badly formed or sized data struct or handing it a duff texture....

After eliminating the possibility of a profile error, I would then drop settings so that detailed architecture is not presented at airports, and test again. If it stops the problem look into the airport scenery as an extended setting can bring items into view lesser settings do not.

Hi Steve,

Textures or objects surely aren't the root cause because the only way I can have a huge impact on reducing these crashes is to reduce my Power Limit setting on my GPU to 70-80%. My SIM settings and content remain entirely unchanged so the exact same textures and bgls etc are still loading but no crash. Mine was entirely random on 100s of tests and tweaking prior to discovering the power limit fix, it wasn't linked to any particular location like Pete is experiencing. It could even happen in tests with the aircraft sitting still on a runway 

9 hours ago, charleslee201 said:

Hi  Pete  have you try  this..., that worked for me 

Browse to and then click the following registry subkey: 
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\GraphicsDrivers
- On the Edit menu, click New, and select the QWORD (64-bit) value from the drop-down menu,
- Type ‘TdrLevel’ as the Name and click Enter,
- Double-click TdrLevel and set the value as 0 (it is set to that by default, but double check) and click OK,
- Close the registry editor and restart your computer for the changes to take effect!

 

 

 

 

These edits have never ever worked for me

[glider1 90]

On 12/7/2018 at 9:53 AM, cj-ibbotson said:

Yes I think the Windows 7 Pro was retail, it came in a large plastic box and cardboard sleeve about book size.  Whilst you were typing your reply I edited my last comment as PSU is working again, the on board lights flickered again when I plugged the cord in.  I then touched the 24 pin connector where it connects to the MB and all MB lights lit up, pressing the power on button system booted up.  Ive been diagnosing cooling issues and moving the AIO cooler about  over the last few days so I may have knocked the power connection though did have power and boot issues just before that I think.  From memory I also might have pulled it out and reconnected it when recently fitting new SSD's and Sata drives and generally reconfiguring drive bays a little.  Its now thrown a spanner in the works regarding the faulty GPU and DXGI crashes as I had started to hope the PSU was the cause.  Its a 1000W OCZ ZX Series Gold 80 I think got in 2011.  The retail version of the 9900K is now £525 on Overclockers where I believe you bought yours so I might just go the whole hog and change as much as I can now, might just order a new PSU as well.  Ive held back as concerned of heat of the 9900K, currently have a Corsair H100i v2 which is 240mm long.  Might buy a 360mm unit H150 i think and a new case

I'm going down the route of thinking it is a power supply problem for myself and possibly you. Your 1000W is getting old at 2011 vintage. The rails might not be as solid as they used to be with caps aging so I'm thinking it is not a bad idea to get a new one.

I swapped out my PSU for a beefier one today and will look forward to testing it over coming days.

The clue for me is that I started getting device hung a lot more frequently a few days ago and I only now realise that at the same time I had added periripherals to the PC to the extent where every single USB port was being used. My VR headset needs USB and when I saw in the event log that the VR driver was doing a user power mode reset only shortly before the device hung came up on P3D, it made me strongly suspect that my PSU rails were dropping off under max load which drops the VR headset, which drops the GPU which P3D reports as a device hung.

Yeah it is an intermittent problem, they are the most difficult to track down.

[cj-ibbotson 77]

20 minutes ago, glider1 said:

I'm going down the route of thinking it is a power supply problem for myself and possibly you. Your 1000W is getting old at 2011 vintage. The rails might not be as solid as they used to be with caps aging so I'm thinking it is not a bad idea to get a new one.

I swapped out my PSU for a beefier one today and will look forward to testing it over coming days.

The clue for me is that I started getting device hung a lot more frequently a few days ago and I only now realise that at the same time I had added periripherals to the PC to the extent where every single USB port was being used. My VR headset needs USB and when I saw in the event log that the VR driver was doing a user power mode reset only shortly before the device hung came up on P3D, it made me strongly suspect that my PSU rails were dropping off under max load which drops the VR headset, which drops the GPU which P3D reports as a device hung.

Yeah it is an intermittent problem, they are the most difficult to track down.

Let me know how you get on. I'd love it to be a simple thing like the PSU though still unsure given my son's 1080 worked fine on Prepar3D when I tested it on my system. Saying that it only needed a single 8 pin power connector and my 1080ti need 8 plus 6 so perhaps struggles to draw the power. Maybe why errors are reduced when I reduce Power Limit setting. On the other hand it passed Time Spy on Extreme without a problem

Chris

[pgde 571]

I tried all of the above (other than modifying the power setting) and nothing worked. So I tried using 3D Mark as described above. Unfortunately I can't run Time Spy since it apparently requires DX12 (as well as a paid upgrade). But I did try running the Firestrike DLC and surprise, the card locked up after only 90 seconds of display. I was running GPU-Z logging and the reading at the time of crash were identical to the P3D card crashes (i.e. spike of GPU usage to 100% twice in a row then lockup). In fact, the crash locked my PC and I needed to reboot.

So, unless anybody has any other suggestions, I will RMA the card early next week. If any of the technical types out there can interpret the GPU-Z log and provide me with some narrative to give to Gigabyte, I would be eternally grateful!

Here is the link to the 3D Mark GPU-Z log: https://www.dropbox.com/s/iszim7elkt07dis/dxgi error log with Firestrike.xlsx?dl=0 -- if anybody has happened to run 3D Mark with Firestrike and done a GPU-Z log, I would love to see it to compare to mine.

Thanks for all your help!!!!

P.

BTW, while I don't have a PSU tester, my BIOS reading (under PC Health Status) show the following:

Rail       Reading

+3.3V      3.363 V

+5          5.160 V

+12       12.240

This doesn't look like a PSU issue?

Edited December 8, 2018 by pgde
Added PSU readings

[cj-ibbotson 77]

10 minutes ago, pgde said:

I tried all of the above (other than modifying the power setting) and nothing worked. So I tried using 3D Mark as described above. Unfortunately I can't run Time Spy since it apparently requires DX12 (as well as a paid upgrade). But I did try running the Firestrike DLC and surprise, the card locked up after only 90 seconds of display. I was running GPU-Z logging and the reading at the time of crash were identical to the P3D card crashes (i.e. spike of GPU usage to 100% twice in a row then lockup). In fact, the crash locked my PC and I needed to reboot.

So, unless anybody has any other suggestions, I will RMA the card early next week. If any of the technical types out there can interpret the GPU-Z log and provide me with some narrative to give to Gigabyte, I would be eternally grateful!

Here is the link to the 3D Mark GPU-Z log: https://www.dropbox.com/s/iszim7elkt07dis/dxgi error log with Firestrike.xlsx?dl=0 -- if anybody has happened to run 3D Mark with Firestrike and done a GPU-Z log, I would love to see it to compare to mine.

Thanks for all your help!!!!

P.

At least you get consistency with crashes as Time Spy ran perfectly well on the extreme stress test and passed. Leaves me thinking then what on earth is causing Prepar3D to crash. My fear is if I go through the very long RMA process with Inno3d (no UK or Europe base so back to Hong Kong) then what if it passes all stress tests with them just like at mine. Will make me think twice and research implications if a product fails.

Chris

[Pete Dowson 558]

2 hours ago, pgde said:

BTW, while I don't have a PSU tester, my BIOS reading (under PC Health Status) show the following:

Rail       Reading

+3.3V      3.363 V

+5          5.160 V

+12       12.240

This doesn't look like a PSU issue?

Not sure whether a static voltage test, without the PSU being under load, tells you much. It's whether it can sustain them under load. Can those values be displayed at such times?

Incidentally, my DXGI problem probably isn't GPU related as it can happen with either of two cards.  I suppose it could be PSU related, but I'm still more inclined to think it is in the area of P3D/Windows/nVidia.

So far it hasn't happened since cleaning the video driver with DDU and installing the latest (471.22, at least the latest a couple of days ago).  But then I haven't had time to do a lot of testing.

Pete

 

Edited December 8, 2018 by Pete Dowson

[Snuffleupagus 327]

A friend of mine just had the same error for him (GTX1060) and same situation as I’m in, it only has happened after he updated to 4.4. Not blaming the 4.4 update, just putting that out there in case it’s relative good info to have. 

Edited December 9, 2018 by pvupilot

[pgde 571]

16 hours ago, Pete Dowson said:

Not sure whether a static voltage test, without the PSU being under load, tells you much. It's whether it can sustain them under load. Can those values be displayed at such times?

Incidentally, my DXGI problem probably isn't GPU related as it can happen with either of two cards.  I suppose it could be PSU related, but I'm still more inclined to think it is in the area of P3D/Windows/nVidia.

So far it hasn't happened since cleaning the video driver with DDU and installing the latest (471.22, at least the latest a couple of days ago).  But then I haven't had time to do a lot of testing.

Pete

 

Good point. I guess I will need to buy a tester for load testing. 

[Guest]

4 hours ago, pgde said:

Good point. I guess I will need to buy a tester for load testing.

I have one ... something to check are the pins in the connectors themselves, sometimes they can get pushed up so contact isn't "strong" between male/female ends.

Cheers, Rob.

[SteveW 1,792]

On ‎12‎/‎8‎/‎2018 at 10:24 AM, cj-ibbotson said:

Hi Steve,

Textures or objects surely aren't the root cause because the only way I can have a huge impact on reducing these crashes is to reduce my Power Limit setting on my GPU to 70-80%. My SIM settings and content remain entirely unchanged so the exact same textures and bgls etc are still loading but no crash. Mine was entirely random on 100s of tests and tweaking prior to discovering the power limit fix, it wasn't linked to any particular location like Pete is experiencing. It could even happen in tests with the aircraft sitting still on a runway 

These edits have never ever worked for me

Still can be I'm afraid - since that is related to the texture rate.

[SteveW 1,792]

On ‎12‎/‎8‎/‎2018 at 11:31 PM, Pete Dowson said:

Not sure whether a static voltage test, without the PSU being under load, tells you much. It's whether it can sustain them under load. Can those values be displayed at such times?

Incidentally, my DXGI problem probably isn't GPU related as it can happen with either of two cards.  I suppose it could be PSU related, but I'm still more inclined to think it is in the area of P3D/Windows/nVidia.

So far it hasn't happened since cleaning the video driver with DDU and installing the latest (471.22, at least the latest a couple of days ago).  But then I haven't had time to do a lot of testing.

Pete

 

So how'd it go then Pete - did you factory default the main profile and create a P3D specific one?

[SteveW 1,792]

The registry item to create is not a fix as I mentioned - it should only be used in testing and debugging certain situations. Remove it in case of recovery with more serious problems.

[SteveW 1,792]

On ‎12‎/‎8‎/‎2018 at 11:31 PM, Pete Dowson said:

Incidentally, my DXGI problem probably isn't GPU related as it can happen with either of two cards.  I suppose it could be PSU related, but I'm still more inclined to think it is in the area of P3D/Windows/nVidia.

 

A lot of games show this error on various systems, it comes and goes, same on AMD and Nvidia cards. It comes and goes because it's usually down to a problem presented in the data heading to it. 

Eliminate the possibility of configuration problems first in NCP. Then eliminate the possibility of an airport item or certain architecture being the cause by reducing the scenery settings.

Edited December 10, 2018 by SteveW

[SteveW 1,792]

...Remember that cleaning and upgrading the driver situation usually does the trick of factory defaulting. However it can still pick up old profile.

On the NVidia site there was a button last week with "sli patch", I didn't go there to see what it was.

[squerble 300]

Pete, do you do anything else on the PC running P3D when this happens? I've noticed my PC is much more likely to give me this error if I'm doing something else on the PC at the same time, be this using Chrome or opening a PDF. I've yet to try a "proper" clean with DDU but sometimes it'll crash as soon as I open a PDF - I'm wondering if it's something to do with P3D losing focus, at least on my end.