Newish build that was fine when built, but few months down the line keeps crashing when gaming

Plastic rivet

Plastic rivet

Plastic rivet

Associate
Joined
17 Oct 2020
Posts
29
Specs:
CPU: AMD Ryzen 9 5900X 3.7 GHz 12-Core Processor
CPU Cooler: Lian Li GALAHAD AIO 360 RGB 69.17 CFM Liquid CPU Cooler
Motherboard: Asus ROG Crosshair VIII Dark Hero ATX AM4 Motherboard
Memory: Corsair Dominator Platinum RGB 32 GB (4 x 8 GB) DDR4-3600 CL18 Memory
Storage: Sabrent Rocket 4.0 2 TB M.2-2280 NVME Solid State Drive
Video Card: EVGA GeForce RTX 3090 24 GB FTW3 ULTRA GAMING Video Card
Case: Lian Li PC-O11 Dynamic ATX Full Tower Case
Power Supply: Corsair RMx (2018) 850 W 80+ Gold Certified Fully Modular ATX Power Supply



When gaming, my computer constantly crashes. Sometimes it's the game that closes down, other times it's a blue screen or a black screen or a complete restart.
Eventviewer shows distributedCOM error sometimes when the game crashes to desktop, which I've tried fixing through enabling permissions but it just keeps appearing with a new source.


When crashing to blue/black screen, sometimes it details a CPU fan error or CPU over temperature error, but when I turn back on, temps are usually fine, around 60-70 degrees at the time usually from gaming, which I believe is well within normal.


I've took CPU heatsink off, checked all wires, re-done thermal paste and installed again. Temps are all over tbh, sometimes idles at 33, other times it can be 40 and at the time of writing this, it's currently at 52 degrees. Yet when I run Cinebench R23, it peaks at 60 degrees and doesn't go any higher, though in some games it reaches 70.

I've ran MemTest 86, which came back with a PASS.

Event view shows a DistributedCOM warning, which I've tried to fix, but just keeps coming back with a different permission.
Reliability viewer shows several hardware errors also for unexpected shutdown, but no idea what's causing this - maybe cpu or cooler.


Any thoughts~?
 
Download hwmonitor and run a benchmark and check temps. Heaven is free and can be looped to stress test the gpu. To test the cpu run blender or cinebench. Do not test hardware using games, to many variables

Have you checked the gpu memory temps? the 3090 has memory chips on the back of the pcb and they can get very very hot which could cause crashs.
 
Haz123 said:
Download hwmonitor and run a benchmark and check temps. Heaven is free and can be looped to stress test the gpu. To test the cpu run blender or cinebench. Do not test hardware using games, to many variables

Have you checked the gpu memory temps? the 3090 has memory chips on the back of the pcb and they can get very very hot which could cause crashs.
Click to expand...


I ran Cinebench r23 and got a 20500 score and hwmonitor shown temps on cpu peaking at 60 degrees.
I ran Heaven and memory temps peaked at 88 degrees.
No crashes from either of them.
 
I would loop Heaven for about an hour and see if the memory temps are still ok but the initial tests are promising.Now to the other issue and thats software. Update bios and reinstall gpu drivers is a good place to start but I assume you have already done that. You do not always need the most upto date bios and drivers and sometimes an older set will be more stable.
What games are you playing that causes the crashes ?
 
Haz123 said:
I would loop Heaven for about an hour and see if the memory temps are still ok but the initial tests are promising.Now to the other issue and thats software. Update bios and reinstall gpu drivers is a good place to start but I assume you have already done that. You do not always need the most upto date bios and drivers and sometimes an older set will be more stable.
What games are you playing that causes the crashes ?
Click to expand...

I'll loop HEAVEN now for an hour whilst I'm on my dinner.

Reinstalled GPU Drivers and Updated BIOS to latest driver as it was quite out of date.

It can be any game tbh. First it started with just games on Steam, so Dishonored 2 in particular. It would crash to desktop and there would be a distributedCOM error in the event viewer, which I tried to fix, but keeps appearing. Then it would eventually crash to black/blue screen or just randomly restart PC. It did this on Warzone, Planet Zoo and Halo. I though it may be the intensity of the games, so I played League and it did the same thing also.
 
Distributed com error is probably a red herring. This sounds like overheating. Have you got good airflow in your case? It may sound daft but are your dust filters all nice and clean?

would be useful to log your temps so you can see the exact temp at the point it happens. If you have a 3090 i'm pretty sure they throttle if temps get too high on the ram or GPU (but you could be past a point of no return if airflow is restricted) so i'd be looking at CPU temps first.
 
At defaults RTX 3090 can dump crazy amount of heat into case.
So could be also other than usual suspects going unstable.

For example memory using Samsung B-die chips are sensitive to temperature.
Cinebench is very bad for that running pretty much inside CPU's caches.
Some Linpack based stress test or Prime would use also main memory and would be good complement for simultaneous GPU stress test.
 
You must log in or register to reply here.
Back
Top Bottom