[SOLVED] 5950X System hard resets at random. No errors in event viewer.

[nunyaabidniss]

Hello all,

I have replaced nearly every part of my build through RMA's because my PC keeps hard resetting randomly. Just kicks off, but the only thing in the event viewer is that it was improperly shut down.

Ryzen 9 5950X, EVGA FTW3 3080 Ti, Gigabyte X570S Aorus Master (F4b BIOS), Samsung 980 Pro 1TB M.2 Drive, 32GB (4x8) G Skill Trident Neo 3600mhz CL14 (2x F4-3600C14D-16GTZNB), Corsair HX1200 PSU.

I have RMA'd the mobo, CPU, New PSU (went from an RM850x), and I replaced the ram (from Flare X 2x16 3200 CL14 to the current ram). I also replaced the AIO thinking it may be a water leak. The GPU works perfectly in another build.

I have contacted AMD and they don't believe it is the CPU, as this is the second 5950X with the same exact symptoms. I upgraded my mobo from an X570 Aorus Elite Wifi to the X570S Aorus Master thinking that may be the issue.

I have tried disabling Global C State, tried a static OC, did fresh installs of WIN 11 through each new troubleshooting step, tried a new M.2, ran the memory without XMP enabled at 2133, updated everything with firmware to its most current version, all of this and I still get random resets. The installation of a new mobo gave me false hope as the system was seemingly stable for around 12 hours but went right back to hard restarting and it did it 3 times in a row.

I installed my known working 3900X in the system prior to the new Mobo being installed and I got random reboots with it as well. I have ran memtest86 and in the middle of the test I get random reboots, but prior to the reboot there are no errors.

My temps never break 80C on both the GPU and CPU.

I am at a loss.

Does anyone have a suggestion for me to look into? I appreciate any help.

 

[Ralston18]

This:

"but the only thing in the event viewer is that it was improperly shut down. "

Also look in Reliability History. Much more user friendly and provides a timeline format that may prove revealing.

Overall, my thought is that power is being lost - just small outages perhaps due to a loose wire or connection.

Or a short somewhere.

Wall outlet, extension cords, power bar, surge protector, power cord to PSU, case switch. Check the electrical path.

Try some gentle wiggling, pulling, and twisting. Determine if the problem is a physical connection issue.

And, as always, be sure that all data is backed up at least 2 x to locations away/off of the problem PC. Verify that the backups are recoverable and readable.

 

[nunyaabidniss]

This:

"but the only thing in the event viewer is that it was improperly shut down. "

Also look in Reliability History. Much more user friendly and provides a timeline format that may prove revealing.

Overall, my thought is that power is being lost - just small outages perhaps due to a loose wire or connection.

Or a short somewhere.

Wall outlet, extension cords, power bar, surge protector, power cord to PSU, case switch. Check the electrical path.

Try some gentle wiggling, pulling, and twisting. Determine if the problem is a physical connection issue.

And, as always, be sure that all data is backed up at least 2 x to locations away/off of the problem PC. Verify that the backups are recoverable and readable.

Thank you for the reply.

I have checked Reliability History as well, forgot to mention that. The only error there is the same thing as in event viewer, improper shutdown.

I rewired the entire build from the ground up twice now, even tried with new cables. I replaced my power cable and my UPS as well.

I thought "ding!" it must be the outlet when I started troubleshooting since it seemed electrical, that wasn't it either.

I have been working on this issue since the end of October actually. I am just lucky I kept my old system.

I am beyond baffled. Also, there's nothing on this PC I can't just redownload. I definitely appreciate that recommendation!

 

[Ralston18]

Baffling indeed....

Take a closer look in Reliability History and Event Viewer.

Try to find a pattern with respect to the "improper shutdowns". Hot, cold, wet, dry, windy, etc...

Consider electrostatic shocks - do you get shocked a lot in your home environment? When the computer is on and audio is being used - do you hear occasional pops and crackles?

Check the case. Use a bright flashlight to inspect all around inside. Especially near the I/O panel. And around all case - switches, LEDs, USB, audio ports, etc..

Make and model router? Check the router's logs (if available and enabled) to see if the router's logs might be catching something.

Check all cables: video, audio, network, USB etc. Swap in known working cables from other systems.

Disable all power savers and sleep modes. Minimize the number of Startup apps in Task Manager.

Worth a look to check into Task Scheduler as well. Some trigger could be involved.

At this point my thought is to pare down to as bare bones a build as you can. If the crashes stop then begin adding things back. Slowly and methodically allowing a day's time between each change/addition.

 

[nunyaabidniss]

Baffling indeed....

Take a closer look in Reliability History and Event Viewer.

Try to find a pattern with respect to the "improper shutdowns". Hot, cold, wet, dry, windy, etc...

Consider electrostatic shocks - do you get shocked a lot in your home environment? When the computer is on and audio is being used - do you hear occasional pops and crackles?

Check the case. Use a bright flashlight to inspect all around inside. Especially near the I/O panel. And around all case - switches, LEDs, USB, audio ports, etc..

Make and model router? Check the router's logs (if available and enabled) to see if the router's logs might be catching something.

Check all cables: video, audio, network, USB etc. Swap in known working cables from other systems.

Disable all power savers and sleep modes. Minimize the number of Startup apps in Task Manager.

Worth a look to check into Task Scheduler as well. Some trigger could be involved.

At this point my thought is to pare down to as bare bones a build as you can. If the crashes stop then begin adding things back. Slowly and methodically allowing a day's time between each change/addition.

I wanted to thank you and give you an update.
The system has been tested on a test bench and I found myself having the same issues. I decided to just start trying voltage and timing settings (again) for the ram and CPU. I ended up switching the cas latency to 15 from 14 (it's 3600Mhz CL14 RAM) and the SOC voltage to 1.15 as well as a few other slight changes. This seemed to somewhat stabilize the system. I started to mess with other voltages and timings and I actually ended up overclocking the RAM to 4000Mhz with a CL of 15 and enabling PBO and setting up Curve Optimizer to a -30. I ran Prime95 for around 11 hours, got no restarts or errors. Ran Cinebench R23 one time after all that and I got a hard reboot. Dropped my OC down a bit to 3800Mhz loosened the timings a bit more to CL16 and dropped the Curve Optimizer from -30 to -25. Ran P95 again all day while I was at work and had zero reboots or WHEA BSODs. Ran Cinebench R23, OCCT and MemTest yesterday and the night before and had zero crashes or errors.

The 5950X and the RAM hate each other unless I manually set voltages, but it seems that once set and tuned I have a stable and much better performing system. If I have one more BSOD or a single hard reset I will RMA both the RAM and the CPU though. For now, I think I will consider this solved but I will update the thread in the future if I find I was premature in my elation.

 

[helper800]

I wanted to thank you and give you an update.
The system has been tested on a test bench and I found myself having the same issues. I decided to just start trying voltage and timing settings (again) for the ram and CPU. I ended up switching the cas latency to 15 from 14 (it's 3600Mhz CL14 RAM) and the SOC voltage to 1.15 as well as a few other slight changes. This seemed to somewhat stabilize the system. I started to mess with other voltages and timings and I actually ended up overclocking the RAM to 4000Mhz with a CL of 15 and enabling PBO and setting up Curve Optimizer to a -30. I ran Prime95 for around 11 hours, got no restarts or errors. Ran Cinebench R23 one time after all that and I got a hard reboot. Dropped my OC down a bit to 3800Mhz loosened the timings a bit more to CL16 and dropped the Curve Optimizer from -30 to -25. Ran P95 again all day while I was at work and had zero reboots or WHEA BSODs. Ran Cinebench R23, OCCT and MemTest yesterday and the night before and had zero crashes or errors.

The 5950X and the RAM hate each other unless I manually set voltages, but it seems that once set and tuned I have a stable and much better performing system. If I have one more BSOD or a single hard reset I will RMA both the RAM and the CPU though. For now, I think I will consider this solved but I will update the thread in the future if I find I was premature in my elation.

I have the 4 x 8 gb cl 14-15-15-34 kit of Trident Z Royals with a 3900x and also had to do some fine tuning to get rid of an occasional pestering BSOD but I ended up with better timings at the same speed. Did you try messing with the SoC voltage? Try 1.05-1.1v. What about the VDDG CCD Voltage? Try 1.0-1.05. VDDG IOD? Try 1.0-1.05. cLDO VDDP? Try 0.950-1.050. Gear down mode set to disabled? BGS set to disabled? BGS alt to enabled? CPU LLC to level 2/3 for the CPU / VDDSOC?