5950x Sudden WHEA errors and reboots when starting games

T

Tazman2

Supreme [H]ardness
Joined
Nov 17, 2000
Messages
5,018
Bought it in July. Specs in sig. Suddenly it just started the other day and noticed in event viewer I have WHEA errors. I'm just using PBO and was using a negative 0.125 offset voltage just fine no issues until now. I notice when I go up ie less negative it starts stabilizing. Should I just try to RMA it at this point or?
 
Or use pbo with curve optimizer and temp limit it via bios and see what happens. That's how I run mine and it's fine with very good speed and temps. Limits at 84C 4.55Ghz average during all core full load boosts to 4.95 during everything else what more could I ask. BTW max boost voltage all core 1.38 single core 1.47. Plenty safe plenty fast.
 
Also, are you using FMAX enhancer or whatever it may be called on your board (0-200mhz boost)? If so that may be your cause. Don't bother with that as it doesn't help performance on Zen 3 chips.
 
GSDragoon said:
Remove the negative voltage offset and go back to default.
Click to expand...
Already tried this last night no change. It has tons of low fps longer before the crash/reboot.

EDIT: set ram and fclk at defaults too same as above.
 
Digital Viper-X- said:
Also, are you using FMAX enhancer or whatever it may be called on your board (0-200mhz boost)? If so that may be your cause. Don't bother with that as it doesn't help performance on Zen 3 chips.
Click to expand...
Not sure will have to check. Its weird cause there was no issues since I bought it till now...
 
Are you using it to mine? I recently replaced my 5900x and the new one does the same, all stability tests passed, I'm beginning to strongly suspect either nbminer or amd graphics drivers. In my case having a mixed nv/amd mining rig doesn't help so I'lll likely move all AMD cards to one PC and NV to the main rig next.
 
Tazman2 said:
Already tried this last night no change. It has tons of low fps longer before the crash/reboot.

EDIT: set ram and fclk at defaults too same as above.
Click to expand...
Check for power supply issues particularly the 12V rail if FPS is erratic and eventually crashes.
 
learners permit said:
Check for power supply issues particularly the 12V rail if FPS is erratic and eventually crashes.
Click to expand...
Most of my errors and along with many folks happen at mostly idle when the PC is just mining and using less than 3% CPU..gaming and other loads including extended P95 and AIDA tests are solid. May be different for the OP though.
 
Check Typical Current Idle / Low Current Idle /// Dust on the cpu pins /// or contaminants such as thermal paste on cpu pins.
 
Last edited:
Did you update your bios recently? Not sure if it's related, but ever since one of the more recent AGESA versions, my 3900X crashes when undervolted a lot more than it used too. I just raised my negative offset and it's fine, without much temperature difference.
 
learners permit said:
Check for power supply issues particularly the 12V rail if FPS is erratic and eventually crashes.
Click to expand...
Also check your RAM with one of the standalone tools that runs overnight.

When i first built my new rig (ASUS X570, AMD 3900X) I would have these weird, random BSODs and shutdowns. Was a RAM issue, which Crucial cheerfully replaced. No issues since, for about 18 months now. Rock solid. (y)
 
undertaker2k8 said:
Most of my errors and along with many folks happen at mostly idle when the PC is just mining and using less than 3% CPU..gaming and other loads including extended P95 and AIDA tests are solid. May be different for the OP though.
Click to expand...
Yeah, this is normally caused by Curve Optimizer having too strong of an offset, but CO can display other rare oddities too. In my case CO would only cause 1 core to crash on rare occasion in near idle instances, but even setting up that specific core with no curve or even a positive one would still cause it to crash; disabling CO altogether, there would be no issues and haven't been for months. I may go back to it in a year with a more mature BIOS, but the benefit isn't currently worth the hassle for me.

edit; I've been told a negative offset in LLC can be more stable, but haven't tested it myself.
 
TheSlySyl said:
Did you update your bios recently? Not sure if it's related, but ever since one of the more recent AGESA versions, my 3900X crashes when undervolted a lot more than it used too. I just raised my negative offset and it's fine, without much temperature difference.
Click to expand...
Not until this started as part of what I could tell AMD I tried, etc. I eventually stuck a 5700G I had laying around from another project and it works fine now no more FPS drops or crashes, etc.
 
It's hard to take away voltage, and then ask the CPU to boost higher. Things stop working properly.
 
Tazman2 said:
Worked for a few months just fine...
Click to expand...
FWIW my first 5800x was fine with pbo +200 for a month or two. Then it wasn't. Moved it to use curve optimizer with a negative offset and it was fine. I sold it and got a different one, and it's much better.
 
I had this happen to me about 2-3 weeks ago and it was a windows update of some sort - it got so unstable it wouldn't even boot, but restore to previous via the troubleshooting menu fixed it - no idea what the update was because I couldn't even get in to windows to see what had been updated recently. It was bizarre, but same exact symptoms, luckily I tried the restore before wasting a ton of time with bios versions or settings, since that episode it's again been rock solid.
 
You must log in or register to reply here.
Back
Top