Nvidia Display Driver Crashing

[Sergster]

My problems started from optimizing the download cache from the warframe launcher.

 

Launched game just fine, stalker popped up in the load screen right before login, boom, kernel driver crash, and then game turns "invisible", I saw my steam, browser page, but mouse was stuck ingame, which ran perfectly fine, closed it via taskmanager and alt-f4 worked too, my card is gtx780 Poseidon, with driver version 361.91 Geforce game ready.

 

No other game causes this issue but warframe.

 

Anyone got ANY idea? This is one way to celebrate +1000 hours used in the game. I really would love to keep playing without having to do any hardware fiddling or whatnot.

[SmokeyJesus]

My problems started from optimizing the download cache from the warframe launcher.

 

Launched game just fine, stalker popped up in the load screen right before login, boom, kernel driver crash, and then game turns "invisible", I saw my steam, browser page, but mouse was stuck ingame, which ran perfectly fine, closed it via taskmanager and alt-f4 worked too, my card is gtx780 Poseidon, with driver version 361.91 Geforce game ready.

 

No other game causes this issue but warframe.

 

Anyone got ANY idea? This is one way to celebrate +1000 hours used in the game. I really would love to keep playing without having to do any hardware fiddling or whatnot.

Not sure if this is related to this exact topic, but I'm having the exact same issue. 560m, Latest driver. Not sure whats happening.

[hahameat]

On 2016/1/6 at 0:51 AM, Cailian said:

Digging up old stuff i know, buuut in case others have this issue (which im sure tons do)..

this crash is a maxwell (should be mostly) card issue (aka 900 series) switching between different power modes (normal clocks, boost speed, energy saving etc)

 

So try going to the control panel and from Manage 3d Settings>Program Settings> from there select "Warframe.exe" > and there change "Power management mode" from "adaptive" to Prefer Maximum performance.

 

Also make sure you're not running any custom over-clocks etc ofc.

 

Especially mid to high and high end will have issues since the game is lets say... weak gfx wise on the card load so these cards are often going in and out of normal to high performance speeds and back constantly.

 

Last but not least, if all of the above didnt do the trick for you, simply use something like MSI afterburner, to Down-clock your graphics card. Simply get your Core Clock to something like -30 or -40 Mhz and it will also help prevent the issue from re-occurring. Just remember to go back and reset to normal speeds when you want to play another game with the normal full speeds.

Thank you very much for help, my computer has suffered several times and the hardware check is normal.I am wondering why it always crashes

[Maj_Brian]

I just started getting this problem today. Recently its made me loose 2 trinity chassis BP lex p bp and Hikou Bp and pouch. Plus about 10 more prime parts. This is really getting annoying and I've been playing for about a year with no problems until now. Now every time I start a mission my screen turns grey or turns off completely and I can only sit helplessly pressing my power button. I have updated my drivers and have a gtx 670 graphics card, any advice on this problem would be fantastic.

I saw somewhere where a DE employee said to turn V-sync on and I have done that and yet no change.

Edited February 27, 2016 by Maj_Brian

[SourVenom]

It seems to be general Nvidia diver / warframe error. Started happening to me as well like a month ago. First i was getting those messeges "Insufficient Video Memory!" after crash, sometimes all went one color (mostly brown or grey) and since last hotfix i get blank screens with game playing ok - i hear sound as normally. Since i upgraded to ewindows 10 yesterday i got that variable sorted out - fresh and different system, fresh drivers, fresh WF install. Yet it still occurs time and time again.

Bonus thing is that Warframe never exceeded 1,100MB of VRAM usage so it couldn't be "insufficient memory" issue since my video card has 4GB VRAM (or 3,5 + 0,5 - depends on how you look at it) ... UNLESS some malfunctioning procedure causes memory leak and floods it at an instant.

Last time it happened to me was on grineer ship (sortie t2 29.02.16), if i recall correctly it happened a few times on Excavation farm on corpus factory tileset on Neptune. But i noticed that it never happened to me when running void missions which is kind of odd because crashes occur mostly after 1-1,5h of playtime and after Saryn Prime pack was introduced i was running missions with team for like 6 hour straight with no crash at all.

 

The reason i have not submitted tickedt yet is that i plan to include all the data i can hoard during such a problem but for some fun reason 4 times i have tried to 'catch' it by monitoring GPU usage via GPU-Z it did not crashed.

Also my playsessions feature heavy alt+tabbing - this may be useful too.

 

After further investigation i can more or less connect it to Nvidia PhysX. I have found a hell lot of THOSE in EE file after crash.

Quote

1879.719 Phys [Error]: GPU particle pipeline failure detected. Skipping particle pipeline execution.
1879.719 Sys [Error]: FATAL APEX error detected. Stopping Apex updates.

.........

1880.015 Phys [Error]: GPU particle pipeline failed, particle data not accessible
1880.024 Gfx [Error]: WorldOnFireEnemySparks failed to allocate vertex buffer data, disabling.
1880.026 Sys [Warning]: WorldOnFireEnemySparks: Reallocating particles, this should only happen after mode resets!

(...)

1895.676 Sys [Warning]: DematerializeParticlesA: Reallocating particles, this should only happen after mode resets!
1895.676 Gfx [Error]: DematerializeParticlesA failed to allocate vertex buffer data, disabling.
1895.676 Sys [Warning]: Radiation: Reallocating particles, this should only happen after mode resets!

I have shortened it a bit - it goes on like this for about 6 thousand lines. 

EDIT 2:

After change of settings according to [revious post in quote below some new info emerged in EE.log

Quote

1603.334 Phys [Error]: physx::apex::CudaModuleScene::destroy: Cuda Error 999
Phys [Info]: -- APEX dump stats begin --
Phys [Info]: Cuda Memory GPU Total allocated 0 Heap size 0 Max allocated 542,638,080
Phys [Info]: Cuda Memory Pinned Host Total allocated 524,288 Heap size 4,194,304 Max allocated 32,550,912
Phys [Info]: Cuda Memory Write Combined Total allocated 0 Heap size 0 Max allocated 0
Phys [Info]: Cuda Memory Host Total allocated 524,288 Heap size 2,097,152 Max allocated 1,048,576
Phys [Info]: Number of static entities in APEX scene 0
Phys [Info]: Number of dynamic entities in APEX scene 0
Phys [Info]: Dynamic update queue length 0
Phys [Info]: Number of particle systems 0
Phys [Info]: Number of turbulence volumes 0
Phys [Info]: Number of attractor systems 0
Phys [Info]: Number of ground emitter systems 0
Phys [Info]: -- APEX dump stats end --

(...) - actually a few iterations of the one above and just before my ALT+F4 action the thing below

 1603.335 Phys [Error]: CUDA error: CUDA_ERROR_UNKNOWN
1603.335 Phys [Error]: CUDA error: CUDA_ERROR_UNKNOWN
1603.335 Phys [Error]: CUDA error: CUDA_ERROR_UNKNOWN
1603.360 Sys [Info]: ResourceLoader::~ResourceLoader() in 0.581ms [Heap: 541,916,256/690,946,048 Footprint: 2,258,477,056]
1603.361 Sys [Info]: ResourceLoader::~ResourceLoader() in 0.193ms [Heap: 541,655,408/690,946,048 Footprint: 2,258,477,056]
1603.361 Sys [Info]: ResourceLoader::~ResourceLoader() in 0.112ms [Heap: 541,379,056/690,946,048 Footprint: 2,258,477,056]
1603.361 Sys [Info]: ResourceLoader::~ResourceLoader() in 0.100ms [Heap: 541,271,904/690,946,048 Footprint: 2,258,477,056]
1603.361 Sys [Info]: ResourceLoader::~ResourceLoader() in 0.188ms [Heap: 541,025,696/690,946,048 Footprint: 2,258,477,056]
1603.362 Sys [Info]: ResourceLoader::~ResourceLoader() in 0.271ms [Heap: 540,756,208/690,946,048 Footprint: 2,258,477,056]
1603.362 Sys [Info]: ResourceLoader::~ResourceLoader() in 0.603ms [Heap: 539,811,664/690,946,048 Footprint: 2,258,477,056]
1603.364 Sys [Info]: ResourceLoader::~ResourceLoader() in 1.52ms [Heap: 537,580,400/690,946,048 Footprint: 2,258,477,056]
1603.365 Sys [Info]: Context shutdown complete [Heap: 534,260,576/690,946,048 Footprint: 2,258,477,056]
1603.365 Input [Error]: InputMgrWin::Shutdown    <<< my ALT+F4

So much for errors, the most heartwarming thing i have seen yesterday was DE_Steves post according to Inaros' 18.5 update

Quote

Physics collision, performance and stability have all been improved, as a long-awaited upgrade of Warframe PhysX libraries has been completed.

Full post can be seen via THIS LINK .

Edited March 3, 2016 by SourVenom
Adding info

[JesterDay]

On 6.1.2016 at 5:51 PM, Cailian said:

Digging up old stuff i know, buuut in case others have this issue (which im sure tons do)..

this crash is a maxwell (should be mostly) card issue (aka 900 series) switching between different power modes (normal clocks, boost speed, energy saving etc)

 

So try going to the control panel and from Manage 3d Settings>Program Settings> from there select "Warframe.exe" > and there change "Power management mode" from "adaptive" to Prefer Maximum performance.

 

Also make sure you're not running any custom over-clocks etc ofc.

 

Especially mid to high and high end will have issues since the game is lets say... weak gfx wise on the card load so these cards are often going in and out of normal to high performance speeds and back constantly.

 

Last but not least, if all of the above didnt do the trick for you, simply use something like MSI afterburner, to Down-clock your graphics card. Simply get your Core Clock to something like -30 or -40 Mhz and it will also help prevent the issue from re-occurring. Just remember to go back and reset to normal speeds when you want to play another game with the normal full speeds.

Thanks a lot! The combination actually done the trick. Not a single blackscreen since.

DE should pay you, saved them a customer.

Best regards

[Virt]

On 1/6/2016 at 11:51 AM, Cailian said:

-Snip-

I tried this, this fixed it for me too! BLESS YOU BRO!

Edit: Nevermind...they crashed after ~2 hours. Considerably better but still, this isn't fair to people who have this bug. Fix this please, DE!

Edited March 6, 2016 by Virtual_Anomaly
Update on fixing issue.