PC Processors

[G1n]

Hello, I bought a new pc a month ago, and since then I haven't able to find a fix for my issue. It's been a real pain and I regret a lot by upgrading from my old FX-8350 just to find so hard to get my system stable out of the box. When playing certain games, mostly on Apex Legends, I would get black screen, sound "stuck in distortion" then after 2 or 3 seconds of black sreen, computer reboot. 
I have updated my bios, chipsted drivers, set everything to auto in bios "load optimized defaults" seems to delay the crashes, but they still happen. Already did fresh windows install "I'm downloading windows 11 to see if it makes any difference" I have disabled xmp and run ram at 3.200 issue is still there. I have tried different Nvidia drivers, replaced PSU. The crash comes with WHEA LOGGER ID 18 error in windows event viewer and KERNEL 41 error. 

System:

MOBO: Asus PRIME 570X-P BIOS:4021
CPU: Ryzen 7 5800X
RAM: G.Skill Ripjaws 3.600mhz "b-die"
PSU: SFlower GX PRO 650W
VGA: ELSA GTX 1660 SUPER
Windows 10 21h2 build 19044
Chipset: 3.10.22.706

[ripcord]

This tech site give some very good tips on what causes WHEA LOGGER Error 18:
Event ID 18: Microsoft-Windows-WHEA-Logger - TechNet Articles - United States (English) - TechNet Wi... 

If you continue to experience WHEA errors, please try the suggestions listed in this thread and here. https://www.amd.com/en/support/kb/faq/ts-tips

I hope this helps you.
Ripcord

[tenzo23]

First check your hardware. The 5800X can get hot pretty quickly and needs a better than average cooler. What are your temps in HWMonitor or HWiNFO while gaming?

I don't know about your board, but most recommend occupying the outermost DIMM slot pair. For example, if you have 2 sticks of RAM, put them in A2 and B2 (or slot 2 and 4, counting from left to right).

Now, even on Auto, your motherboard may be trying to push too much boost causing the WHEA errors.

I just found this thread on the Asus ROG forums and it seems like that's the case - https://rog.asus.com/forum/showthread.php?123946-x570-Prime-not-respecting-PBO2-limits

Looks like there are two different menus to adjust Precision Boost and your board has some options that mine doesn't.

But, I'd try going to that AiTweaker menu, Disable Fmax Enhancer, and the rest Auto.

I'll post pics of my 5800X PBO settings you can try.

[G1n]

Thank you, I'll check that thread and yes I'm looking some budget but good cooling. Temps while gaming hit 89ºC in two cores, just like two times in 5 hours, the rest won't pass 79ºC. 

Yes I would like to try some settings you could share, I will disable Fmax Enhancer and report back.

[tenzo23]

So with my 5800X and MB, I found the sweet spot for Manual PBO Limits was 180 PPT, 180 TDC, and 120 EDC.

Then I started with -20 Curve Optimizer on all cores and ended up being stable with -20 on my 2 fastest cores and -30 on the rest.

Then I raised Max Boost in 25Mhz steps and ran Cinebench23 each time. Boost at 125 and 150 gave me the best scores and temps. At 175+ I started to get WHEA errors.

[G1n]

Thank you, I'll try these setting as soon as possible. When I tested gaming with everything on auto, temps were quite high; but crash delayed by a lot. Then I tried some PBO limits to   PPT = 120, TDC = 75 EDC = 110 with no CO. I tested temps and they were quite lower, however Voltage with no CO and CO -15 all cores seems to give me the same voltage under load "testing with CoreCycler-v0.8.2.4" core by core, I get almost 100% constant 1.45v while running this test, it ended showing no error on any core with CO -15 all cores, but I see same voltage as when CO was disabled. I read that constant 1.4v+ = fast degradation, is this safe? I checked while playing Apex and I saw 1.29v super low and it crashed very fast. With CoreCycler mostly 1.45v but no error.... Why games crash but most test just pass fine. I was forced to revert CO to disabled so Apex won't crash in just minutes. As far as for power limits, I'm scared of trying yours because they look waay higher than amd defaults, and mine with PBO set to AUTO = instant 90ºC in most stuff. So imagine setting it higher....

[G1n]

I did a fresh windows 11 install "was using windows 10" and found this https://www.reddit.com/r/Amd/comments/relgkv/amd_chipset_31022706_windows_10/
I had installed the same " 3.10.22.706" chipset driver from Asus, and some users reported there that it was causing issues for them, like reboots. So I went ahead and installed "3.10.08.506" From AMD website this time. Furthermore I went into bios and disabled PBO2 "Since with windows 10 testing any setting or the most soft touch to CO will result in faster crash while gaming" Keept Fmax disabled, XMP enabled and the rest all AUTO. 

Got quite higher voltage while installing games from Steam and using browser and moving around Windows. But I'll test this setup gaming now and report back. So far I got different voltages while not using heavy software or gaming, compared to when I was using newer chipset from Asus and not older one from AMD like I'm using now. Let's see if it makes any difference with the WHEA LOGGER errors and with the crashes while gaming.

I was gaming and AVERAGE CPU Core Voltage was "1.43" seems absurd high.... Everything on AUTO and PBO2 disabled. No crashes so far, but IDK if this chip is gonna degrade if I play games a lot....

[tenzo23]

Don't be concerned. 1.43V is normal and these Zen 3 chips safely go up to 1.5V at times.

[elstaci]

From looking at your PC's PSU unit it seems to be a low budget Chinese made PSU that most likely is causing your Reboot issues during gaming when the GPU and CPU are using a lot power.

Download and run OCCT PSU Test and see if it fails or reboots as soon as you start the test. If the test run long enough make sure you check the PSU Outputs of 12/5.0/3/3 Vdc. They all should be within 5% tolerance.

Check mark those three PSU output voltages to be monitored on OCCT.

Also keep an eye on the Temperatures and Fan RPMs when running the PSU Test. ALso run both the CPU and GPU tests to see if it reboots.

[Diptenkrom]

What chipset is your Mobo? i have that same chip on an Asus X570 Tuf, and it has been flawless. I would agree that the PSU could be an issue, also you need a good heatsink/cooler. get the PSU tested and probably get a name brand 80+gold or better PSU.

[elstaci]

I also have a Asus TUF Gaming X570 Pro (WiFi) with the same BIOS version as you have on your Asus Motherboard plus the same AMD CHIPSET driver set which fixes the Windows 11 flaw concerning AMD processors without any issues at all.

[G1n]

It is not the PSU I know that, and I know about the thermal range this was fixed by just setting PBO limits a small amount below optimized defaults; so for now I don't need that "Even tho I actually need that" as my CPU is running cold. All I know about the chipset is the x570 I don't know any more about that.

[G1n · Accepted solution]

After intense testing playing games, system was stable only after setting the following Bios values:

PBO2: DISABLED

PBO1: MANUAL : PBO LIMITS : PPT 130 TDC 85 EDC 130
PBO1: FMAX ENHANCER: DISABLED
PBO1: SCALAR MANUAL: X2

DIGIPOWER CONTROL:

VDDCR CPU Load Line calibration: HIGH
VDDCR CPU OFFSET MODE SIGN: [-] 
VDDCR CPU OFFSET VOLTAGE: 0.07500

POWER SUPPLY IDLE CONTROL: TYPICAL CURRENT IDLE

Enabling PBO2 "PBO under AMD OVERCLOCKING" with all set to AUTO results in games crashing and WHEA LOGGER id 18 related errors. 
I just want to believe I got a garbage quality chip compared to others, and that it may just not want to run with the PBO2 set to AUTO "boosting to 4,85"

[elstaci]

I would try and see which BIOS setting exactly was causing your crashes.

In another words start one by one enabling the BIOS settings that you changed and see when you PC starts crashing again.

Wait one to three days and if no crashes then change another back to default again and so on.

I would start with the POWER SUPPLY IDLE CONTROL: TYPICAL CURRENT IDLE  since inexpensive PSU have trouble with this setting.  Actually "Typical Current IDLE" is recommended because many PSUs can't handle the first settings.

If you computer starts to crash with Power Supply Idle control back to the way it was then at least you know for sure that setting is one reason for the crashes. Then I would leave it as "Typical Current Idle".

Then I would start one by one putting the other BIOS  settings back to default and see which one again causes crashes.

Like I mentioned wait a few days to make sure you PC is stable when you change one BIOS setting back to the way it was before or on default.

But I just mentioned the above because that way you can find out exactly which BIOS Settings(s) was  causing your PC to be unstable.

Good troubleshooting by the way.

[G1n]

I tried and it was not the Power supply option, in fact I reverted it to "AUTO" which now let my voltage go below 1.1 on IDLE. It was at 1.2xV on IDLE with it set to typical current idle; so after realising it had nothing to do with my crashes now I enjoy even lower voltage on IDLE.  Setting optimized defaults will result in whea logger and crashes while playing, unless I disable PBO2 and set manual values it seems the only way to prevent this from happening. And the PSU is just fine, I have seen expensive PSU do crap performance, so it has nothing to do with price but more with what offer a specific model, no matter the price. I bought this Sflower Legion GX pro Gold 650W just because two friends of mine bought it as well and they run end hardware on it and have not got any issues at all. I ran CoreOptimizer the tool, and my CPU got Silver badge at silicon quality, explains a lot; it's close to bronze so it may run like trash in AUTO settings "trying to behave like gold platinum CPU's" make it clown time trying to achieve 4,85. So I think AUTO settings just overvolt it trying to force a bronze/silver chip achieve what better luck chips do on less voltage.

[dhump0]

i have gigabyte x570 gaming x and i am having the problem from last 3 months since i have purhcased the build. finaly a week ago i have fixed the issue. by settings PBO setting to motherboard and curve optimiser to negative 25 on all cores. Settings core clock to 4.7 ghz and max voltage to 1.4. and the chipset driver is older version from 2021 works fine for me on windows 10 now. but on windows 11 it was not much stable for now i am sticking on windows 10. kernel 41 error and WHQL error both are gone for now.

[G1n]

I don't know where did you get that info but this PSU even for budget price, is all go quality. I already tried in heavier systems and they run just fine. In fact, I just bought it to test, I had an EVGA 600W that ran old rig for 8+years and it was still crashing. 
I already ran OCCT one hour free test and came with 0 errors. I ran CoreCycler for 4 hours and no errors on cores, ran many cinebench r23 and they all end just fine. It is not related to the PSU, that's for sure.  I will try the OCCT PSU test just to please doubters. And also comment the possible fix I have managed to put into work "so far good scores".

[Diptenkrom]

what kind of idle temps are you getting? what GPU?

[G1n]

Idle temps with current bios settings are 35C-45C and the VGA is a GTX 1660 Super. 

[BisonActual]

I have msi b550 mag tomahawk, msi gaming geforce 3080 10g, ryzen 7 5800x, ek 420mm aio, Samsung 980 m.2, seasonic focus gx-850 +gold full modular, x2 g.skill ripjaws 16g, windows 11 pro. I game for sometime 20 minutes and my computer will reboot. No BSoD's, just a full reboot. I reapplied my thermal paste, deselect the hybrid mode on my cpu, I run the game at medium graphics, ran a Furmax test with no issues, and I have did a clean install of windows 11. I ram a mem test, I have NZXT to watch my Temps and they haven't gone over 74c. I have checked all my drivers, and I'm at a loss.

[Thomasdk1405]

Hi

Glad that you solved it - had same issue but solved it differently:

https://community.amd.com/t5/processors/5800x-crashes-now-solved-most-likely-for-all-ryzen-5000/m-p/...

Load optimized defaults and enable xmp / DOCP - save and boot and enter bios again.

Go to advanced menu - AMD overclocking/precision boost overdrive/pbo - this setting is at auto - change to advanced.

In new row (PBO limits) change to motherboard and hit F10 (save and boot) and nothing else. Not sure it is like this but in one menu it is cpu doing overclock and in the other it is the motherboard - when you change this to motherboard only one is doing the overclock. Anyway main thing is to stop this crashing !