Question 7900 XTX dual monitors with different outputs potentially causing timeouts?

[faultD]

Built my current PC in November after my GTX 1080 finally died. I decided to give AMD a shot for my new GPU and went all out on a new rig. 9800X3D CPU, AIO cooler, the works.

My current monitor situation is using an ASUS VG278 as my primary and a BenQ XL 2411Z as my secondary. (I used to use the BenQ as my primary but since it only supports HDMI and DVI, and the 144hz won't work when using the HDMI and the XFX Merc 310 7900 XTX doesn't have DVI, it's been relegated. I now have a DP connection to the ASUS and an HDMI connection to the BenQ)

It started out getting GPU driver timeouts when I would have a twitch stream or youtube video on my second monitor while playing a game, mostly CS2, but has timed out in other games like Marvel Rivals and God of War, but has been completely fine on less intensive games like HOI4 (even with intensive mods like Kaiserreich) and PowerWash Simulator. I have tried changing my GPU clock speeds, CPU speeds, messed around with XMP/EXPO, did complete driver reinstalls (minimal, driver only, complete AMD Adrenalin), made sure my power draw was sufficient, monitoring my temps, reseated the GPU, and did a complete reinstall of drivers using DDU.

I found that disconnecting my secondary monitor from my PC solved my issues with timeouts in CS2. I can't be sure if this is an issue with different outputs, or just the BenQ monitor getting old (it's backlight is burnt out and because I'm not in the states they won't send me a new one, plus the model seems to be out of production).
Looking for similar experiences and possible fixes aside from buying a new monitor with DP input.

 

[Ralston18]

PSU: make, model, wattage, age, condition (original to build, new, refurbished, used)?

Any error codes, warnings, or even informational events being logged by Reliabiiity History/Monitor or Event Viewer just before or at the time of the "time-outs"?

 

[faultD]

PSU: make, model, wattage, age, condition (original to build, new, refurbished, used)?

Any error codes, warnings, or even informational events being logged by Reliabiiity History/Monitor or Event Viewer just before or at the time of the "time-outs"?

be quiet! Pure Power 12 M 1000W, had it since November, brand new

This is what Reliability Viewer consistently puts out for all of these timeouts

 

[Ralston18]

141 could be a hardware failure. But not necessarily so.

However, there are some things you can do

Double check all drivers. Ensure that DDU did indeed find and install the correct drivers.

For the most part download drivers directly from the manufacturer's website, manually install and configure.

Ensure that the source website is truly that of the manufacturer. Just because the manufacturer's name appears in the path/URL does not mean that the website is that of the manufacturer.

No third party tools or installers.

= = = =

Run "dism" and "sfc /scannow" to find and fix corrupted or buggy files.

= = = =

Power down, unplug, open the case.

Carefully clean out dust and debris.

Verify by sight and feel that all cards, connectors, RAM, jumpers, and case connections are fully and firmly in place.

Use a bright flashlight to look for signs of damage: bare conductor showing, melting, pinched or kinked wires, browned or blackened areas, corrosion, cracks, scratches, loose or missing screws, swollen components, dead bugs, moisture....

Refer to the motherboard's User Guide/Manual and replace the CMOS battery as a matter of elimination.

Hopefully all that happened is the something has simply come a bit loose due to expansion/contraction as the system heats up and then cools. Or perhaps just vibrations over time.

 

[faultD]

141 could be a hardware failure. But not necessarily so.

However, there are some things you can do

Double check all drivers. Ensure that DDU did indeed find and install the correct drivers.

For the most part download drivers directly from the manufacturer's website, manually install and configure.

Ensure that the source website is truly that of the manufacturer. Just because the manufacturer's name appears in the path/URL does not mean that the website is that of the manufacturer.

No third party tools or installers.

= = = =

Run "dism" and "sfc /scannow" to find and fix corrupted or buggy files.

= = = =

Power down, unplug, open the case.

Carefully clean out dust and debris.

Verify by sight and feel that all cards, connectors, RAM, jumpers, and case connections are fully and firmly in place.

Use a bright flashlight to look for signs of damage: bare conductor showing, melting, pinched or kinked wires, browned or blackened areas, corrosion, cracks, scratches, loose or missing screws, swollen components, dead bugs, moisture....

Refer to the motherboard's User Guide/Manual and replace the CMOS battery as a matter of elimination.

Hopefully all that happened is the something has simply come a bit loose due to expansion/contraction as the system heats up and then cools. Or perhaps just vibrations over time.

Right after I posted the first time, I went and made sure all the connections were good. I reseated the GPU, PCIe cables, and reinserted the 20+4 ATX cable because I thought maybe the +4 section could not be engaged.
Then I ran SFC /scannow and got a BSoD DRIVER_POWER_STATE_FAILURE and it rebooted. did SFC again and it completed successfully.
I tried a game of CS2 Deathmatch without any crashes but when I did a ranked game, I had multiple crashes, this time with 117 instead of 141, and some mid-game graphical glitches, but a quick rejoin of the game fixed the glitches.
After that, I restarted the game and got almost to the end of another ranked game before getting a crash, but this time it went back to being 141. Once the game ended, I decided to try a Factory Reset + reinstall of Adrenalin, but got the BSoD DRIVER_POWER_STATE_FAILURE again. Rebooted and was able to reinstall Adrenalin and Factory Reset, but this hasn't solved it either.
Tried DISM /Online /Cleanup-Image /CheckHealth and it completed within the second successfully.

Also tried with a roommate's monitor as the secondary monitor, connected with DP. Worked perfectly fine, no crashes.

Gonna see if this can be cause for an RMA.

 

[Ralston18]

Make and model monitors:

1) Combination and connectivity with errors/crashes?

2) Combination and connectivity without errors/crashes?

Some monitors have drivers - check accordingly.

 

[faultD]

Make and model monitors:

1) Combination and connectivity with errors/crashes?

2) Combination and connectivity without errors/crashes?

Some monitors have drivers - check accordingly.

ASUS VG278 (main monitor DP), BenQ XL 2411Z (secondary monitor HDMI), Lenovo ThinkVision (roommate's monitor DP)

With crashes: Any combination when using DP and HDMI from the GPU on two monitors

Without crashes: Single DP monitor, Double DP monitors

All drivers in device manager for monitors are up to date, no drivers found for any of the monitors on the manufacturers' websites

Threw a ticket to XFX to see if they have anything to say about this, will post if they have anything of relevance.

 

[Ralston18]

Interesting.

My suggestion is to use Task Manager and Resource Monitor to observe system performance.

Use both tools but only one tool at time.

Leave the tool window open and viewable while experimenting with various monitor and connection combinations.

The objective being to discover some difference in the system's use of resources between no crashes and crashes.

 

[faultD]

Interesting.

My suggestion is to use Task Manager and Resource Monitor to observe system performance.

Use both tools but only one tool at time.

Leave the tool window open and viewable while experimenting with various monitor and connection combinations.

The objective being to discover some difference in the system's use of resources between no crashes and crashes.

Tried this, whenever the crashes happen it obviously sharply declines the GPU clock speed, VRAM, etc. Nothing outside of that tho, CPU is fine, non GPU fans are fine, cooler is fine.

 

[Ralston18]

Possibly related to HDMI.... ?

Are you using any video adapters?

Use Process Explorer to learn more about what may be happening.

Microsoft, free.

https://learn.microsoft.com/en-us/sysinternals/downloads/process-explorer

May be a bit tricky at first to get a sense of how to view and sort the presented processes.

Look for the prescence or abscence of some process that corresponds to the decline in GPU clock speed.

Or some change in the resource(s) being used by one or more processes.