Question Screen going black, peripherals turning off, yellow DRAM light on motherboard

[shadowdn]

Hello all, I recently built a new system with the following specs:

• 7800X3D
• MSI MAG B650 TOMAHAWK
• 32 GB DDR5 6000MHZ
• RTX 4070ti Super
• Corsair RM750x

The system will run flawlessly for days at a time, sometimes weeks, but every once in a while (almost randomly) my screens will go black, the lights on my peripherals will turn off and the yellow DRAM light on my motherboard will be a solid yellow. The rest of the RGB inside the system itself stays on and the fans spin but the only way I can get it to recover is to hold down the power button and reboot.

This seems to be worse on certain games namely Chivalry 2 and Metro Exodus (meaning it happens more frequently).

I have re-seated the RAM, updated the BIOS, and ran a memtest that returned 0 errors.

I have run cinebench and 3dmark and both not only run as expected but the temperatures are great. I have EXPO and Game boost turned on in the BIOS as well as a -0.1v undervolt on the cpu. Has anyone ran into this before or have any idea what might be causing it?

 

[Ralston18]

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

History of heavy use for gaming, video editing, or even bit mining....?

Disk drive(s): make, model, capacity, how full?

Look in Reliability History/Monitor for any error codes, warnings or informational events captured just before or at the time of the screens going black and/or the yellow DRAM light.

 

[shadowdn]

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

History of heavy use for gaming, video editing, or even bit mining....?

Disk drive(s): make, model, capacity, how full?

Look in Reliability History/Monitor for any error codes, warnings or informational events captured just before or at the time of the screens going black and/or the yellow DRAM light.

PSU: Corsair RM750X, brand new for the build
Usage: Gaming and general browsing
Boot drive: Teamgroup MP33 512GB, 1/3rd full
Gaming drives: WD 1 TB SN570 x 2, half way full

The only errors I can find at the time of the issue is an "ASUS com service shutting down" error and a "windows has shut down unexpectedly" error.

 

[35below0]

Assuming you have an Asus GPU, this hints at a GPU problem. But it could also be a faulty PSU if it's not delivering enough power for the 4070 Ti. It needs 285w minimum, so it's possible the power spikes overwhelm the PSU.
The fact Chivalry and Metro Exodus suffer more issues seem to further point in the direction of the GPU.

Try disabling it and run on integrated graphics. See if you get any similar behavior.
Alternatively try with another GPU, if you have one lying around.

Also, you should try installing a more powerful PSU. At least 850w. It's possible you will have to buy an 850w or more PSU, but i wouldn't spend money unless/until i was sure that was the solution.

 

[shadowdn]

Assuming you have an Asus GPU, this hints at a GPU problem. But it could also be a faulty PSU if it's not delivering enough power for the 4070 Ti. It needs 285w minimum, so it's possible the power spikes overwhelm the PSU.
The fact Chivalry and Metro Exodus suffer more issues seem to further point in the direction of the GPU.

Try disabling it and run on integrated graphics. See if you get any similar behavior.
Alternatively try with another GPU, if you have one lying around.

Also, you should try installing a more powerful PSU. At least 850w. It's possible you will have to buy an 850w or more PSU, but i wouldn't spend money unless/until i was sure that was the solution.

Cyberpunk, horizon forbidden west, witcher etc... all run without this issue appearing which are pretty intensive games, definitely more than chivalry and metro in some cases. So I dont think that really points towards anything, I was just calling out the randomness. Also I can run those stress tests all day long and the pc never hiccups and they are undoubtedly more straining then games. It really is a weird problem and I cant reproduce it whatsoever, just happens randomly with days in between events. 750W is plenty for the system.

 

[35below0]

I agree that Horizon forbidden west is more demanding, but we don't know what games do internaly that causes these problems.

750w is not as much as you think. Your GPU can demand a lot more than it's normal power draw in short bursts.

I again suggest you disable the discreete card and run on iGPU to test whether issues are still there. Of course performance will be completely different on iGPU but it's worth trying to figure out if there's something wrong with your GPU.

 

[Ralston18]

This:

"The only errors I can find at the time of the issue is an "ASUS com service shutting down" error and a "windows has shut down unexpectedly" error."

Any error code numbers?

Also look in Event Viewer. Requires more time and effort to navigate and understand but may find other issues.

= = = =

And I am not a fan of those stress tests. They have some degree of value but overall not real world.

Real stress tests will cause problems and maybe break things. No manufacturer is going to produce a stress test that could lead to an end user's computer being damaged. And even if so, there are probably some disclaimers (i.e. "use at your own risk", "not responsible for", or other such caveats).

As above try the iGPU etc..

 

[shadowdn]

This:

"The only errors I can find at the time of the issue is an "ASUS com service shutting down" error and a "windows has shut down unexpectedly" error."

Any error code numbers?

Also look in Event Viewer. Requires more time and effort to navigate and understand but may find other issues.

= = = =

And I am not a fan of those stress tests. They have some degree of value but overall not real world.

Real stress tests will cause problems and maybe break things. No manufacturer is going to produce a stress test that could lead to an end user's computer being damaged. And even if so, there are probably some disclaimers (i.e. "use at your own risk", "not responsible for", or other such caveats).

As above try the iGPU etc..

The reliability monitor doesn't give me a code number unfortunately and event viewer is so saturated with events I'll have to wait for it to happen again and note the time so I can look closer. As for running the iGPU or a different graphics card.....like I said initially this crash happens very sporadically and sometimes wont happen for days and days. Since I don't have a way to reproduce the issue reliably I don't really want to run without my GPU for potentially days on end. I should note - the crashes most often happened when saving/exiting/loading instead of during gameplay itself, although there were one or two that happened while I was playing.

 

[shadowdn]

So I had this happen 4 times tonight over the course of 25 minutes playing chivalry. This was the error each time in event viewer:

Faulting application name: atkexComSvc.exe, version: 1.0.0.1, time stamp: 0x6156aa76
Faulting module name: atkexComSvc.exe, version: 1.0.0.1, time stamp: 0x6156aa76

After deleting Armoury Crate and all of its associated plugins, about ~9 pieces of software, I was able to play the rest of the night with no issues at all. Note - I played Helldivers 2 for about 1.5 hours prior to this and it ran flawlessly, so something with Chivalry, and the other games in the past was upsetting that software for some reason.

 

[Ralston18]

FYI:

https://www.partitionwizard.com/partitionmagic/atkexcomsvc.html

 

[shadowdn]

FYI:

https://www.partitionwizard.com/partitionmagic/atkexcomsvc.html

I dont have the ASUS Motherboard utility in my programs list. Either it was deleted along with all the armoury crate stuff or I never had it in the first place...which makes sense since I have an MSI mbo. Either way I'll see if it continues to happen or not after deleting all that stuff.

 

[COLGeek]

I dont have the ASUS Motherboard utility in my programs list. Either it was deleted along with all the armoury crate stuff or I never had it in the first place...which makes sense since I have an MSI mbo. Either way I'll see if it continues to happen or not after deleting all that stuff.

Sounds like remnants of the application are still present, at least that is how this scenario seems to be playing out.

Is it same to assume that when you built this system, you recycled a previous installation of Windows (vice a fresh, clean install)?

Or, you inadvertently installed the Asus app by mistake?

One of these scenarios would explain its presence.

 

[shadowdn]

Sounds like remnants of the application are still present, at least that is how this scenario seems to be playing out.

Is it same to assume that when you built this system, you recycled a previous installation of Windows (vice a fresh, clean install)?

Or, you inadvertently installed the Asus app by mistake?

One of these scenarios would explain its presence.

It was a fresh install of windows off a USB boot drive. Second option is more likely, Armoury Crate installed a bunch of extras. I hope that was the cause otherwise I'm back at square one.

 

[COLGeek]

Honestly, a new fresh install might be your best bet. That will ensure all the AC stuff is completely gone.

 

[shadowdn]

Well after a while of the issue being gone it happened again playing Manor Lords a few weeks back. I reverted my under volt on the CPU and it hasn't happened Since. Thinking that may have been the root cause.

 

[faalin]

After deleting Armoury Crate and all of its associated plugins, about ~9 pieces of software, I was able to play the rest of the night with no issues at all. Note - I played Helldivers 2 for about 1.5 hours prior to this and it ran flawlessly, so something with Chivalry, and the other games in the past was upsetting that software for some reason.

Is your GPU from Asus? Did you do a fresh windows install for this computer or move an existing OS from your last build over. Not sure where you are getting the Asus Armoury Crate from as is should not have been installed.

At this point i would do a fresh windows install. Move any data you want over to a second hard drive or an external device. Remove all drives except the OS drive, format, and reinstall. If your GPU is from Asus do not install anything for it, go to Nvidia's webpage and download the driver for it and install.

I would also recommend a new PSU as the one you have is not ATX3.0 compliant. Here is a good read on ATX 3.0 vs 2.0.

While yes on paper your ATX 2.0 750w PSU meets the requirements, it does not meet the new 3.0 standards that motherboard and GPU vendors are now following and allowing these power spikes to happen.