PC Processors

[aetheralG]

Build:

CPU: 5800x (Noctua D15 cooler) - Max temp 64 degrees under load

GPU: Radeon 6800xt (Sapphire Reference Model) - Max temp 78 degrees under load

PSU: Corsair RM850x (850 watt)

RAM: Corsair Vengeance LPX 3200mhz

MOBO: Gigabyte x570 Aorus Elite Wifi

Issue: My new build crashes when under gaming loads and the pc reboots itself (no BSOD). At first, i thought the problem was due to insufficient power so I swapped my 750w power supply with a brand new 850w power supply (that's 200 dollars I'm not getting back due to return policies). I've tried reinstalling chipset drivers, graphical drivers, disabling XMPs, disabling SAM, disabling boosting, and plugging my PC into the wall outlet directly (instead of into a surge protector). I have run windows memory diagnostic, as well as 1 pass of memtest86 (will do more overnight), and I have tried several bios revisions (only ones that support Zen 3).

All my temperatures in HWInfo and all seem perfectly fine even upon crashing.

The only thing has seemed to work is running the 5800x as a 6 core processor using the BIOS settings. This causes the system to become stable. I'm extremely confused as to why this works, but it does.

Do I just have a faulty 5800x and need to RMA? Has anyone else encountered this problem? I'm very confused.

[simaem]

Hey Bud 
You and many others are having this same issue like me with different processors, please check some threads here.

[mstfbsrn980]

You need to upgrade your BIOS. If you are still having problems, you need to increase the CPU core voltage with the BIOS. If you're still having problems, you need to reduce your CPU speed with CPU ration with the BIOS. If you still have problems, return your motherboard. 

Restart problems usually indicate that the motherboard is faulty.

EDIT: You should use the BIOS in default settings, and also remove tools such as Ryzen Master from your system.

[katatonia]

I have the same issue, can't use processor unless i either disable all but one core, or i undervolt my processor to 1.3v, and then still it crashes under heavy load.

I tried with two different, both brand new, motherboards with latest bios. They are not the issue, the CPU is.

[mstfbsrn980]

Specs for the 5800x are available in this picture.

The system I am currently using does not contain an AMD component. But I managed to solve some stabilization problems that I had before by doing some changes with the BIOS for an AMD system.

Looking at the specs for the 5800x, CPU Core Ratio seems to be 38.
So... Would you go to the BIOS and try to give the CPU Core Ratio value 36?

Also, 1.3vcore value can be quite low. When you increase this value, does the error count decrease? 

Obviously, minor BIOS changes for a newly purchased system may make the system very stable...

[URV05]

I am also encountering problems with a 5800x and a b550m aorus pro motherboard.

My issue is that the system shut downs mostly when I'm restarting  windows or arbitrary under various types of loads.

First I tough the issue is the psu, swapped and it' the same. Then I replaced the ram with a pair from the QVL and it was the same with and without XMP.

My temperatures under load are fine, and i can stress the cpu at 100% without issues. Even at 80-90 degrees it didn't shuts down, however It does it randomly even when idle or booting cold.

Reinstalled windows several times, installed latest bios and chipset drivers and removed my nvme ssd and installed windows on a ssd from the supported list...reseated everything on the MB and it was the same.

What seems to be stable for a few hours now is the fact that i disabled Core Performance Boost.

Now I am sitting at 3.8 without any boost without encountering any shut down so far...

I am starting to think that the CPU is not quite fine, or the motherboard bios is the issue.

[Jazzinplayer]

Having the same Issue. Also a x570 Aorus Elite motherboard. Just uploaded a new post with the whole explanation and temporary solution.  

Disable Core Boost Performance from Bios and run at base clock speed. You can enable XMP and it will work fine. 

I think motherboard is not being able to correctly manage voltages under auto settings, this should be Bios/driver related

---

[Jazzinplayer]

Exactly what happens to me. Deactivate CPB and run the CPU at base clock, the activate XMP on you DDR4. That will work stable and cause no crashes. Is not a desirable fix but at least is stable and somehow good performance

---

@URV05 wrote:

I am also encountering problems with a 5800x and a b550m aorus pro motherboard.

My issue is that the system shut downs mostly when I'm restarting  windows or arbitrary under various types of loads.

First I tough the issue is the psu, swapped and it' the same. Then I replaced the ram with a pair from the QVL and it was the same with and without XMP.

My temperatures under load are fine, and i can stress the cpu at 100% without issues. Even at 80-90 degrees it didn't shuts down, however It does it randomly even when idle or booting cold.

Reinstalled windows several times, installed latest bios and chipset drivers and removed my nvme ssd and installed windows on a ssd from the supported list...reseated everything on the MB and it was the same.

What seems to be stable for a few hours now is the fact that i disabled Core Performance Boost.

Now I am sitting at 3.8 without any boost without encountering any shut down so far...

I am starting to think that the CPU is not quite fine, or the motherboard bios is the issue.

---

[aetheralG]

I have tried this and it does not solve my particular problem... the only stable solution remains to run the processor at 6 cores only unfortunately. Disabling Core performance boost does not help in my case.

[aetheralG]

Unfortunately, I've tried disabling Core performance boost and have not achieved stability

[mrsense]

What motherboard and bios version?

[aetheralG]

Gigabyte Aorus Elite x570 gaming Wifi. BIOS F31J

[Nekela]

Is there anyone having this issue on a different motherboard brand? Everyone I have seen having a problem is on a gigabyte board

[tim716]

Yes. I have a 5900x on an ASUS ROG Strix X570-e and having the same WHEA BSODs and self-reboots. 
I never have them under load but only AFTER a heavy load. Like I stopped a game or a burn-in and in several seconds there is a BSOD. 
Sometimes (seldom) it can happen 2-5 seconds after starting a game.

And yes, CBP disable helped. But of course I really hope for a normal solution from AMD and/or ASUS. Using the brand new CPU at half power is frustrating.

[Nekela]

Hey Tim, any chance of being able to see what the voltage is doing when it is coming off of a heavy load? It sounds like it’s undervolting aggressively when coming off boost and causing the system to become unstable. 

[tim716]

Hey! 

If you can tell me the utility to use (will AIDA64 do?) and which voltages to check then I can check this.

[Nekela]

If you can consistently reproduce the crash in aida64 that will be perfect. Watch the voltage across all cores and see if any dip below 1.04 or go above 1.4 when the crash happens

[tim716]

Hmm...Thank you that you made me look there. 
I didn’t even have to reproduce the crash. I see only one core voltage parameter. But in idle it jumps from 0.2v (!!!) to 1.0v, sometimes even up to 1.4 for a moment. Under all cores burn in the voltage is steady 1.0v +-0.02.

PBO And Core Boost are off, so this is even with stable system settings. 
I can reproduce the crash turning Core Boost to auto, but what the sense if the voltage jumps like crazy even now?

What is this crap and is it possible to do something about it?

[Nekela]

Try going in to the bios and setting the performance enhancer to level 3. Then go to the PBO settings and set PBO to enabled and Fmax enhancer to enabled. After that go to the advanced tab and open AMD overclocking. Select PBO here and set PBO to advanced and the limit to motherboard. Also make sure that eco mode is turned off

[tim716]

Thank you really much! 

Unfortunately, it hasn't solved the problem entirely, the system still BSODs or reboots 15-20 seconds after the CPU goes from a burn-in load to idle most of the time with your settings (and CPB on). 

And the core voltages of 0.2V in idle are still there :(

But it hasn't crashed yet after exiting games, i.e. in real applications. So it looks like the system is more stable with these settings. I think I'll leave it like that for now and test further. 

Is there maybe a way to hard limit the lower core voltage, or some other way to try and make the system stable?

[kumarab123]

Had the same problem on a new build. 5600x and B550 Gaming Plus, 1660s and 32Gig DD4 3200Mhz. Developed 3 days after being totally stable. Random restarts, it even started crashing in the BIOS (so that ruled out windows/driver related issues). At first I thought it was a PSU issue, but the same day, within 10-12 hours the motherboard just stopped breathing, no post at all, all debug LED's lit up solid white. Bought all stuff locally from a vendor so the RMA was good and the mb was replaced within 2 hours. It's been running fine for 3 days now (touches wood).

What I'd suggest is to boot into the BIOS. Just let it stay there on the hardware monitor page. See what it does. If it crashes, it's a 99pc chance it's the motherboard, or a 1pc chance it's a PSU(or some other power related issue). Although not impossible, it's very unlikely the CPU itself is a lemon.

Regards

[tim716]

With great help from some Russian gurus I finally found (I hope) a solution for my case.

Just to remind you, my problem specifics were crashes in transitions from loads to idle, in idle within 30 seconds after a transition, or when applying a load again after these 30 seconds. NEVER under a load. Turning the Core Performance Boost off eliminated the issue together with the CPU performance.  If you're in the same boat, try this, it should help.

The following is for the ASUS bios, for other vendors the same parameters may be hidden in another place.

The system is stable so far with the following BIOS settings:

Go to AMD overclocking, set the Presicion Boost Overdrive to Manual. Some additional parameters will appear. In there:

1. (The main thing) Set the EDC current limit to 200A.

2. (Just in case) Set the power limit to 130W.

3. (Just in case) Set the temperature limit to 83C.

1 is an increase, 2 and 3 is a decrease. Leave at zeros all the rest there.

Also, just in case, set Idle Voltage to Typical, Global C-states control to Disable, check that ECO mode is Off. Then you can set Core Precision Boost back to On, everything should work.

Looks like the MB and its BIOS wasn't tested with a 5000 CPU at all (or, if it was, it was like "Ok, it boots, that means it works, great, the job's done), and the BIOS just doesn't know about the larger peak currents of Rysen 5000s, and the BIOS' "digital fuse" is just too small for a new CPU. When changing its clocks the CPU, probably, tries to draw more current, the "fuse" (EDC current limit) kicks in and the CPU malfunctions and produces a BSOD.

These currents (or how the "fuse" works) also definitely depend on the MB and/or the CPU heating (I didn't have any BSODs when cooling the open case with a cold hair fan), that explains why not everyone with the config like mine has the same problem, people with better cooling (or a colder GPU) might be ok at defaults.

[rumple]

This looks like a great thing for a youtube video.  =)  Good job.  I think you may have nailed it.

---------------------------------- My edit of Tim716's Advice: -----------------------------------------------

Background:

BIOS just doesn't know about the larger peak currents of Rysen 5000s, and the BIOS' "digital fuse" is just too small for a new CPU. When changing its clocks the CPU, probably, tries to draw more current, the "fuse" (EDC current limit) kicks in and the CPU malfunctions and produces a BSOD.

These currents (or how the "fuse" works) also definitely depend on the MB and/or the CPU heating (I didn't have any BSODs when cooling the open case with a cold hair fan), that explains why not everyone with the config like mine has the same problem, people with better cooling (or a colder GPU) might be ok at defaults.

---------------------------

THE SOLUTION:

Go into BIOS. Disable CBP, save, reboot. Go back into BIOS.
Set the performance enhancer to level 3.
Then go to the PBO settings and
set PBO to enabled
set Fmax enhancer to enabled.
Open the advanced tab and open AMD overclocking.
Select PBO here
set PBO to advanced and the limit to motherboard.
(The main thing) EDC - Set Spike VRM-out current limit to 200A. (later we can back this down, probably AGESA 1.2)
(Just in case) PPT - Set the socket power limit to 130W. (depends on mobo)
(Just in case) TDC - Set the vrm thermal limit to 85. (depends on mobo and proc)
EDC is a temporary massive increase, PPT here a decrease. TDC depends on your cooling of VRMS and such.

Leave at zeros all the rest in that menu
Set Idle Voltage to Typical
Set Global C-states control to Disable
MAKE SURE THAT ECO MODE IS OFF.
NOW you can reboot, go into BIOS and set Core Precision Boost back to On, everything should work.

[Luca2490]

Off chance you would know similar instructions for a gigabyte board I'm on an X470 Ultra gaming

[RailgunKDA]

Can't find "performance enhancer 3" or fmax" on my x570 aorus elite bio's. Any help would be greatly appreciated.

[Namredel]

Made an account just to say thanks it appears this helped my issue out or it hasn't happend in the 3-4 hours gaming session I did and it usually happens in that time frame. I thought it was thermals at first but it wasn't. I have the Asus crosshair hero VIII if anyone wanted to know. Hopefully they get a bios update to fix this? I long for the day I can use my 5800 to it's full power.

[r0gues4int]

You think this may work for me?

I have the same problems with a 5500 XT, Cpu is 3 3100 and a Gigabyte mobo A520M DSH3

---

@tim716 wrote:

With great help from some Russian gurus I finally found (I hope) a solution for my case.

Just to remind you, my problem specifics were crashes in transitions from loads to idle, in idle within 30 seconds after a transition, or when applying a load again after these 30 seconds. NEVER under a load. Turning the Core Performance Boost off eliminated the issue together with the CPU performance.  If you're in the same boat, try this, it should help.

The following is for the ASUS bios, for other vendors the same parameters may be hidden in another place.

The system is stable so far with the following BIOS settings:

Go to AMD overclocking, set the Presicion Boost Overdrive to Manual. Some additional parameters will appear. In there:

1. (The main thing) Set the EDC current limit to 200A.

2. (Just in case) Set the power limit to 130W.

3. (Just in case) Set the temperature limit to 83C.

1 is an increase, 2 and 3 is a decrease. Leave at zeros all the rest there.

Also, just in case, set Idle Voltage to Typical, Global C-states control to Disable, check that ECO mode is Off. Then you can set Core Precision Boost back to On, everything should work.

Looks like the MB and its BIOS wasn't tested with a 5000 CPU at all (or, if it was, it was like "Ok, it boots, that means it works, great, the job's done), and the BIOS just doesn't know about the larger peak currents of Rysen 5000s, and the BIOS' "digital fuse" is just too small for a new CPU. When changing its clocks the CPU, probably, tries to draw more current, the "fuse" (EDC current limit) kicks in and the CPU malfunctions and produces a BSOD.

These currents (or how the "fuse" works) also definitely depend on the MB and/or the CPU heating (I didn't have any BSODs when cooling the open case with a cold hair fan), that explains why not everyone with the config like mine has the same problem, people with better cooling (or a colder GPU) might be ok at defaults.

[___jedi___]

First - thank you and Russian guru for that settings, that stops games from crashing in 2 minutes.

Now I can play for hours, but still crash with WHEA Error 18 remains.

Wandering if anything else can be tuned in BIOS? Mobo is MSI X570 Tomahawk WI-FI, latest beta BIOS.
I've used following:

- PPT 200W
- EDC 333A
- TDC 333A

[rumple]

These values matter a lot, and I can't promise these are the right settings for your CPU and mobo:

Set VDDP - 960 mV (to max of 1.06)   
Set VDDG IOD - 1010 mV (to max of 1060mv)
Set VDDG CCD - 940 mV (to max of 1.1v)

Uncertain about ProcODT.  (Possibly 40 to 45)

Set SOC voltage of 1.15
Set SOC LLC 3 (I think that's what it's called)

Disable Resizable BAR support
Disable C-States
Set PCI_E1 Gen Switch = Gen 4
Set Chipset Gen Switch = Gen 4

Set RAM voltage manually, based on XMP data read from RAM using a utility or whatever (try 1.42v) .    Optional: Set RAM parameters, manually -   Future:  Set GearDownMode disabled 

--------------------- The following is not really as important: 
IF you have the feature PBO2: Within PBO Curve Optimizer: Set voltage magnitude adjustment for all cores to a positive value offset between 7 - 10. 
               Can lower down as low as -10 (negative ten) if system remains fully functional during burn in testing.

Temporary: Within PBO advanced settings, Set EDC limit to 200  (might bypass hard locks due to bug. Can set back when new AGESA BIOS updates fixes)

Temporary: Disable PBO AND at the same time, set CPU voltage from AUTO to OFFSET mode and add about 0.05v ! The goal is to target about 1.36 volts at idle.

----------------------------------------------------------------------

[Joumbax]

Rumple answer saved me.

My problem is not exactly the same.

Ryzen 5 3600 with Trident z neo 16-19-19-39 3600.

My problem is that I wasn't able to bring the fabric clock, whatever the voltage combinaison I tried, and I tried a lot.

The settings you provided in the first part are the only one that allowed me to bring the Flck at 1800.

No one ever talked about SOC LLC but it seems to be the settings that changed the game for me !!!

I created an account explicitly to Thank You, you're the real MVP !

[rumple]

Yes, you can try the various PBO settings, in addition to the default setting "Enabled".  On my motherboard, the "enabled" setting works, but it sets the PPT, TDC, EDC to values that I consider too high.  When I choose "Enhanced" 1 - 4 , the values are a little better on my mobo.  But I really like the option called "Advanced" - "set to motherboard limits".  Because, in theory, the manufacturers know better than me.  Finally, if you still crash, choose "Eco mode 95w" temporarily, and see if you still crash.  I was able to bypass a bug in my system by using the following values:  PPT 122, TDC 83, EDC 220. EDC might be on the high side, the defaults should be PPT=140, TDC=85, EDC=142.  But I think they use 200 because of a temporary glitch. And I think PPT should be locked lower than default during testing to isolate a bad chip.

Now having said all that.  hard crashing under load can also be your RAM or power supply.  In my case, I had too or too little  ram voltage and after the chips heated up under load, the RAM errors glitched the OS.  So I went into BIOS and set my ram to the factory settings, 1.35v, and instead of using the XMP profile, I manually input the first 5 parameters: in my case, CL16-16-16-36-50-etc.  And sure enough, after I got the voltage right, and set the parameters instead of using XMP, everything booted.  Sometimes you have to clear the old settings that get autosaved in the "AMD Overclocking" menu (area in BIOS separate from the normal settings), just set the parameters in the mobo OC section instead of the autosaved section.  After the system was stable, I was able to slowly add voltage to the RAM, and by the time I got to 1.46v I was able to use tight timings of CL-14-14-14-34-44 , roughly.  And that sped up some of my operations by 10%.  The total system benefit is at least 2%.

[devvv4ever]

Thank you for all your tips.

Just registered here to say the one and only thing that helped was to replace the CPU with another one. Sent back the faulty 5800X to the vendor, got another one from a different vendor with a different charge number (00066 was faulty while 00063 working - but I don't know if this has something to do with it).

All these problems are probably coming from a bad QA from AMD and thus they are selling CPUs that are not working within the specifications.

* BIOS updates did stabilize a little bit - maybe due to adjusted voltages. (MB: Aorus x570 Ultra with F33a BIOS)
* The curve adjusted (to about +15) helped most, but a brand new 460 EUR CPU should not needed to be tweaked or scaled down anyway.
* A new power supply didn't change anything (replaced it just to be sure).
* The new CPU works out-of-the-box super-nicely with top performance without any crash neither in Windows nor in Linux.

My recommendation: Immideately replace your CPU if you have unstable systems like WHEA crashes.

Cheers,
Bernhard

[Jimi]

This helped me fix crashes in game with an Asrock B550m and 5600x. Almost gave up after trying many different fixes that had worked for me in previous builds. I had not even begun to overclock. Thank you!

[Destroyn]

For my ASUS B550-F (Wifi) and 5800x with AIO all I needed was a BIOS and chipset update from 8/2020 to 1/2021 to fix full system crashing in Apex Legends ready up screen (gaming).

This is with 30 mins of testing, so try other fixes too. 

[six]

Yes i am having this issue while playing WarZone, ASUS TUF X570 PLUS, 5800x, Gskill 3600 32gb, RX 6800.

[devvv4ever]

Simply replace the CPU, it's faulty and you will have no more issues at all. Trust me - mine is now working more than 6 weeks without ANY issue after I replaced the CPU. The rest of the system is identical.

[six]

I was hoping for a solution, i see a few in this thread but nothing that seems like a fix id like to implement (gimping performance).. I have started up a return and replace process. luckily it seems like my cpu is in stock at ccl so i should get one pretty quick. 

[devvv4ever]

I wish you luck. I ordered mine from a different vendor, just to be sure that the chance getting a very different one would be higher.

[vishal9195]

Same issue with my 5600x 

Motherboard is B550 AORUS PRO AC

AS i disable core performance boost everything works fine no bsod.

Works fine with manual cpu clock ratio 46 and v core voltage 1.35 but somehow CORE PERFORMANCE BOOST is making it crash.

[PurpleClouds0503]

So...

i have had the same problem.

My PC was crashing while playing certain games. The screen just became black und the PC turned on again.

Strangely enough I had no problems with any benchmarks or certain other games.

Windows was also running stable.

After a crash Windows event log showed ...

-----------------------

WHEA Error - Event 18 - serious hardware error

Reported from component: CPU-Core

Source of Failure: Machine Check Exception

Type of failure: Bus/Interconnect Error

Core-APIC-ID:  2

-----------------------

My hardware ...

CPU: Ryzen 7 5800x

Mainboard: ASRock x570 Extreme 4

-----------------------

I fixed the problem by updating the bios/ uefi from v3.00/ AGESA 1.0.8.0 (this was the first version suppporting ryzen 5000 processors) to v3.60/ AGESA 1.2.0.0.

I hope this helps someone..