Random Game Crashing - May 2020

[Rex Havoc]

I have recently returned to the game after a several month absence. Previously, I had no problems with running the game, but now I am constantly experiencing game crashes. This has happened on various maps with different characters and under various conditions (combat, standing still, running down the street, the character select screen, etc). I have no new hardware or software changes (other than windows and driver updates) since my last time playing.

 

In an effort to stop the crashing, I uninstalled the game, cleaned the registry with CCleaner, restarted the machine, and reinstalled the game. The problem still exists. I believe the issue to be graphics card/driver related (see Event Viewer info below), so I rolled my NVIDIA driver back from version 445.87, release date: 04/15/2020, to the previous version (442.74, release date: March 19, 2020). The problem remained.  I have reduced the graphics settings to their lowest (something I've never had to do before while playing) and it does not stop the random crashing.  I also have ran several troubleshooters (NVIDIA, Microsoft, etc) on the video card and they all come back good.

 

I haven't experienced this issue with any other applications or games. Specifically, I have tried to recreate the graphics crashes with Fallout 76, World of Warcraft, Warhammer Return of Reckoning (private server) and various windows applications. So far it only happens in City of Heroes.  I'm at a loss of what's going on.  I opened a support ticket, but wanted to share my story here incase anyone has any insight or suggestions.  

 

Here are my computer's stats:

 

Windows 10 64bit

Intel Processor i7-4930k [email protected]

NVIDIA GeForce GTX 780 video card

28 GB RAM

 

Here are several separate post-crash entries from Event Viewer:

 

1. A TDR has been detected.

The application must close.

Error code: 7

(pid=2960 tid=15164 cityofheroes.exe 64bit)

 

2. Unable to communicate with the display driver.  The application must close.

Error code: 2

(pid=6584 tid=5052 cityofheroes.exe 64bit)

 

3. Faulting application name: cityofheroes.exe, version: 1.0.1327.64, time stamp: 0x5e8b0c59

Faulting module name: nvoglv64.dll, version: 26.21.14.4274, time stamp: 0x5e6bf217

Exception code: 0xc0000409

Fault offset: 0x0000000000fb9b79

Faulting process id: 0x3364

Faulting application start time: 0x01d625ac5fc2aa09

Faulting application path: D:\Games\CoH\hc-bin64\cityofheroes.exe

Faulting module path: C:\WINDOWS\System32\DriverStore\FileRepository\nv_dispi.inf_amd64_ab1f86d7947528e0\nvoglv64.dll

Report Id: 1c1f2ef3-b40b-4516-89c5-ab8bd5396f11

Faulting package full name:

Faulting package-relative application ID:

 

The game is pretty much currently unplayable. Which sucks because I was really enjoying myself. Thanks for taking a look at this and any suggestions you might have to fix my issue.

[GM Tock]

TDR up above refers to "Timeout Detection & Recovery".  In short, the Graphics Card stopped responding within a timeframe (usually 2 seconds) when Windows was expecting a response from it.

 

NVidia's advice on TDR errors is one of the two directions -- either lengthen or disable Windows' TDR feature (not recommended, as TDR saves components from fatal errors and eventual malfunction), or reduce the workload you're giving to the card.  TDR is usually indicative of an issue with the hardware itself.  It's not necessarily a death sentence, but it does mean the card has reached a limit or found problems.  Some possibilities:

• Driver Issues.  This is the easiest of the three to consider.  And you just walked through it above.
• Heat Damage.  The GTX 780 isn't exactly waiting for the scrap heap, but it's not exactly a spring chicken anymore.  Any performance optimizations intended to eke more out of the card should be walked back, such as overclocking, turning off post-processing effects in Homecoming that are GPU heavy (such as Bloom, Depth of Field, Ambient Occlusion, etc.) and reducing workload (if you have multiple monitors, turn one off and see if it persists.  Every monitor over 1 is additional Video RAM in use that is drawing extra energy inside the card.)  Also, checking for good airflow inside your case may help: check and replace damaged fans, make sure it's dusted, or if the graphics card is showing a bit of age, consider a replacement.
• Power Supply Issues.  TDR issues also happen when a Power Supply is failing to deliver voltage to the Graphics Card, or worse, is close to failing.  A remedy is replacing the power supply for your computer.  Its possible as well for a lead to fail on a PCI-E connector, where a second one works fine.

 

Heat Damage and Power Supply issues with laptops becomes fairly monolithic for folks, as neither the DC intake (the plug and brick is replaceable, the DC Jack may not necessarily be) nor the Graphics Card chips are usually replaceable past professional service or a visit to iFixIt.com, $35-100 in parts and a lot of patience.  That shortens the list further, because there's no easy recourse if the components are indeed failing short of replacing the machine as a whole.

 

Not great news, and possibly some work to do.  I hope it works out for you.

Edited May 9, 2020 by GM Tahquitz

[Rex Havoc]

Thanks for your reply!  I cleaned out the case and verified that all fans were functioning. The heat monitor on the case reads in the green (66F to 75F at various locations; motherboard, video card, etc).  I'll have to look into the power supply and/or lead issue.  I'll also see if I can grab a loaner graphics card from a friend to swap out with the GTX 780 and see what that results in.

 

The one thing that is odd, is the issue is only with City of Heroes. I have tried to recreate it in other games (graphic intensive areas, zoning from one area to another, etc.) but haven't been able to recreate the issue. The games I have tried so far are Fallout 76, World of Warcraft, and Warhammer Return of Reckoning.

 

edit: I forgot to mention this is a desktop and not a laptop.

Edited May 9, 2020 by Rex Havoc

[Amaterasu]

It's happening with my RTX 2070 too, full custom wc loop so it's not temps, they never get close to thermal limits.

It's a driver fault, that's what system events says.

[GM Tock]

That advice is specific to TDR errors, which may not necessarily apply in your situation.