Question GPU stability Issues ?

[LiquidDew]

Problem (TLDR):
Over the past few months, I've been experiencing severe, intermittent frame drops that often lead to full application crashes when running any 3D application (mainly games). I'm looking for advice on how to improve stability and prevent these frequent crashes.
My GPU is an MSI GTX 1080 Aero OC from 2016 (ancient I know).

Specific Examples:

• Overwatch: boots up and handles menus flawlessly, but usually freezes and crashes two to three seconds into character selection. When it crashes, I receive a system message saying 'render device lost.'.
• World of Warcraft: Severe stutters after logging in to a character for the first minute or two, sometimes resulting in a crash.
• Satisfactory: Freezes and crashes in the main menu / after loading in.
• GTFO: Freezes and crashes immediately, or upon loading in to a level.
• Destiny 2: crashes in the loading screen before character select.
• Factorio: 0 issues ever.

Every time I have checked the Event Viewer after a crash the event log has had an entry from "nviddmkm" with an event id of 153. The message is less than helpful:

The description for Event ID 153 from source nvlddmkm cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer.

My Thoughts:
The common thread for problematic games is that they all require 3d rendering. The frustrating thing is that, although crashing is a regular occurrence, the vast majority of the time it occurs within a few seconds of starting the game or loading in to the world. If the game doesn't almost immediately crash, I can game for essentially indefinitely without having issues. I have noticed sometimes tabbing out of the game can cause the stuttering to restart (although i only remember this happening in WoW).

Attempted Fixes:
I have replaced much of the other hardware since these issues started (Mobo and PSU). I have also done a fresh install of windows, ensuring GPU drivers are up to date. I am very convinced this is a GPU issue. Since issues vary from application to application, I will list the performance in various games. I have tried modifying the TdrDelay value in the registry, and after tweaking that extensively it seems to only make crashing more regular.

Question:
Considering the current GPU market and that I can almost always get into a game eventually (or at the very least play Factorio!), I can't justify upgrading the card atm. Is there any advice that anyone can give me to allow myself to get the last bit of mileage out of my GPU?

System:
CPU: i7-9600k
GPU: MSI Aero 1080 OC
RAM: 2x16 Trident Z DDR4
Mobo: ASUS Prime H310M-A R2.0
PSU: Corsair SF600
Hard Drive: Samsung 960 Evo 1TB
OS: Windows 11 Home, Build 26100

 

[Lutfij]

Welcome to the forums, newcomer!

PSU: Corsair SF600
How old is the PSU?

ensuring GPU drivers are up to date.
What driver version are you working with?

As for your GPU, what sort of temps are you seeing for said GPU?

If I were in your shoes and I had to narrow the GPU as the source of the issue, I'd drop said GPU into a known working platform, with ample power at the entire system's disposal and see if the issue persists. If it doesn't then the issue might very well be your motherboard(it's BIOS). Speaking of motherboard, what BIOS version are you on for said motherboard?

 

[LiquidDew]

Welcome to the forums, newcomer!

Thanks! I thought I had an account here from years ago but apparently not!

How old is the PSU?

It is less than a year old. I replaced it when i started having issues.

What driver version are you working with?

I am running the 580.88 driver according to the Nvidia software, Task Manager says 32.0.15.8088. Weird that those don't match up Anyway, freshly installed as of a few days ago.

As for your GPU, what sort of temps are you seeing for said GPU?

At idle I'm looking at temps of 45 C, as I duplicated the issue currently the temperature was 71 C at the time of the crash.

Speaking of motherboard, what BIOS version are you on for said motherboard?

I am currently running a bios version 1004.

I first had these issues almost a year ago with a different motherboard, PSU, and chassis. I forget the details, but I was having the same GPU issues alongside system stability issues. After an especially gnarly system crash (felt more like a power outage than anything) I suspected my PSU, as well as the riser cables my GPU was installed with. I RMAed the PSU and got a new chassis and motherboard: the riser cable needed for my chassis, so that had to go, and i suspected my motherboard might have been damaged in the crash.

I'd drop said GPU into a known working platform, with ample power at the entire system's disposal and see if the issue persists.

I can't do this immediately but it is a good idea. I'll be in the area of a friend with a working PC in a few weeks for a wedding, I'll just pack my GPU in my bags! I'll have to make sure that he has a decent PSU to rule out power issues though. I believe he has a 750 though so it shouldn't be a big deal.

If I was having power issues does it make sense that just the GPU would be having trouble? I always thought that if a system was power starved it would just fully crash.

Thanks for the prompt response and for helping me out!

 

[SkyNetRising]

Is there any advice that anyone can give me to allow myself to get the last bit of mileage out of my GPU?

Try lowering GPU power limit in MSI Afterburner.
This may improve gpu stability.