cancel
Showing results for 
Search instead for 
Did you mean: 

Processors

gontos
Adept I

Re: Ryzen 5900x: System constantly crashing/restarting WHEA-Logger ID 18 and critical error Kernel-P

Jump to solution

My solution was replacing the cpu, after that, no problems whatsoever...

0 Likes
Cmdr-ZiN
Challenger

Re: Ryzen 5900x: System constantly crashing/restarting WHEA-Logger ID 18 and critical error Kernel-P

Jump to solution

@Soulsa It's all been covered many times. It works for some, not everyone. However it's a work around not a fix.

Ryzen with infinity cache just can be unstable with various combination of components and narrowing them down is difficult because for most of the time it's stable.

I heard a story of 2 friends they built identical systems, one had the issue the other didn't. They swap CPU's with each other to see if the issue followed the CPU. However after they swapped CPU's, neither of them had the issue again.

I also heard Ryzen 3800X's had the issue and it was eventually fixed for most by a BIOS update.

For others it could be fixed by a CPU, GPU, PSU, MOBO or RAM swap, I've even heard of a sound card causing it.

I've thought I've had it fixed multiple times only to have it reappear months later. I still want to get to the bottom of it, however in all the systems I've ever owned, both AMD and Intel, I've never had an issue like this.

Soulsa
Adept I

Re: Ryzen 5900x: System constantly crashing/restarting WHEA-Logger ID 18 and critical error Kernel-P

Jump to solution

It sucks, I'm over the moon AMD are doing fairly well BUT whenever I have had Intel they always just worked....NVIDIA also.... My 5700xt has had DRIVER problems for over 6 months, AMD KNOW also and we bombard them about it all.

 

I can't help but think it's always the driver team letting down the products...

0 Likes
Cmdr-ZiN
Challenger

Re: Ryzen 5900x: System constantly crashing/restarting WHEA-Logger ID 18 and critical error Kernel-P

Jump to solution

@Soulsa they have a vastly smaller team so it's hard to compete but their software is maturing.

Infinity fabric is new tech, never really been done before on a home PC. Intel is getting into similar tech maybe things will be less stable these days.

Maybe they're just pushing it too far. Maybe they made an assumption about the combos of rigs people would make.

Without knowing the cause it would be hard to say. I do suspect it's a combination of issues though.

kneel420
Adept I

Re: Ryzen 5900x: System constantly crashing/restarting WHEA-Logger ID 18 and critical error Kernel-P

Jump to solution

So....

After all the hoop jumping.....  It looks like once I upgraded my BIOS for my Gigabyte X570 Aorus motherboard from F30 to F34, no more reboots and WHEA-Logger 18 errors.   Mostly this issue I think is a combination of hardware and firmware version incompatibility.  

Cmdr-ZiN
Challenger

Re: Ryzen 5900x: System constantly crashing/restarting WHEA-Logger ID 18 and critical error Kernel-P

Jump to solution

@kneel420 After replacing my CPU I couldn't reproduce the issue as much as tried. A couple of months later it returned. Everything was fully updated when I replaced the CPU a couple of months ago and because it was working I didn't change anything except auto updates in Windows.

My current BIOS is 2401 the lastest is 2420 - AGESA V2 PI 1.2.0.3 Patch C this is a few revisions later, I'm gonna try it next and see if it helps too. It probably is a combination of few things for most people. Maybe the chipset driver changed and needs a BIOS update or something.

Maybe it's my RAM G.Skill F4-3600C16D-32GTZNC, I could of sworn it was on the ASUS Mobo QVL before I bought it, but after noticing the MOBO wasn't on G.Skills QVL page, I  noticed on the ASUS MOBO page it was now only showing for the 3000 series. The MOBO is now back on the G.Skill QVL page, I was hoping this is because the latest BIOS. Still not on ASUS' page yet.

Anyway it's weird that RAM is going off and on the QVL lists.

 

kneel420
Adept I

Re: Ryzen 5900x: System constantly crashing/restarting WHEA-Logger ID 18 and critical error Kernel-P

Jump to solution

Yeah its definitely combinations of hardware and firmware version incompatibilities.   Really comes down to ensuring all your drivers are up to to date.  

My system is rock solid now and was about to replace the processor or worse. 

Would have found out sooner but Gigabyte's site was broken due to alleged ransomware attack so many of their files (bios, drivers) are 404ing still.  

Be careful with the advice online, just go through ALL your hardware make sure drivers updated it should solve any issues.  I thought I had done that but Gigabytes BIOS tool kept telling me it was updating the bios firmware but it wasn't, probably due to that issue.  I had to do it manually

0 Likes
Cmdr-ZiN
Challenger

Re: Ryzen 5900x: System constantly crashing/restarting WHEA-Logger ID 18 and critical error Kernel-P

Jump to solution

@kneel420 I've always kept my system fully upnto date, except for the time after I swapped my CPU. While testing my new CPU that was the only change I made, I ran the PC constantly for 17 days straight trying to make it fail and couldn't.

The issue returned a couple of months later however.

If this was just software related It would of been fixed the whole time during the 6 months I was trying to fix it.

The instabilities in Ryzen are just not acceptable.

0 Likes
eliotcole
Adept II

Re: Ryzen 5900x: System constantly crashing/restarting WHEA-Logger ID 18 and critical error Kernel-P

Jump to solution

I'm going to make two post, apologies in advance. This post was actually from the 10th, the next will be me catching up with y'all.

----------

This was a long post but it's not letting me post, so if AMD ever do look at these forums, that's the reason for the multiple posts.

Anyone reading ... you can skip to your reply if you don't really care that much (I reply email style, always have, always will) ... but mostly everything I'm writing is general musings.

Most importantly, though.

To Everyone In The Thread - CAN YOU REPRODUCE THE FAILURE?

It's going to be really helpful if y'all can isolate a specific thing (or things) that causes the crashes here. Because if these start to overlap, then anyone that does eventually try to do something *real* about this will be able to take that and run with it, diving into all the code and components that are involved.

For me, this is quite reproducable:

  • Time-Spy - Resolution 1440p, Rendering at 4K, 60FPS, DX12 (and 11) - It happens as the avatar/thief turns away from the giant back into the museum. The only time anything reaches high utilisation/temps during this is at the very start, then it's buttery, otherwise. No spikes on anything at time of crash.
  • Total War: Warhammer II - Highest everything, GPU memory unlimited/not-unlimited, various other settings - It happens directly after a successful siege battle with Grom in one of the immediate territories. When the post battle screen is on, and you click to continue after selecting your bounties, boom, power down. Yes, I'm aware this game engine is one of the worst optimised pieces of crap out there. I'm also aware that it's one of the best RTS ever made (I've even said so!), plus the fact that it's badly optimised makes it a good test.

I'll happily try any other tests that others have confirmed cause the issue, although ZiN's recent relapse has be doing permanent wearyface.:-(




ZiN (and everyone, tbf), I'm aware that your issue has come back, and I'll be coming back to that eventually, but thanks for the reply, none-the-less.

My PC is back now from SCAN (they've been great, if a bit uncommunicative) ... and what I'll be doing is:

  1. Putting all the drives except the spinner in the PC
  2. Not connecting any extraneous devices apart from the monitor. What I can remember that's:
    • The HDMI to AVR (4K res) that I mostly use for the surround outs to the Razer headset.
    • The USB 3 switch (unpowered).
    • External drives.
    • Second USB 3 switch (powered).
  3. Booting up with the new BIOS that (yes, I could've done it) SCAN put on and reported zero issues apparently testing at the specs I run on.
  4. Sighing that something.
  5. Testing immediately.
  6. I expect it to fail, which would indicate (for me) driver/software conflict. If it doesn't, then great ... if it does ...
  7. I'll have to reluctantly upgrade this (my main) install of Windows and try again.

I'll then install linux, and test there.


@Cmdr-ZiNwrote:

 for some reason I don't get notified of thread updates anymore, but I did see the mention thanks.

... Good luck I hope you solve it.




 

Cheers for this, Anzu.


@Anzu34wrote:

 
- CPU : AMD Ryzen 5900x 
- MB : MSI B550 Tomahawk
- GPU : MSI Geforce RTX 3080 Ti Trio Gaming
- AIO : NZXT Kaken x73 360mm RGB
- RAM : Crucial Ballistix 2x16 Go 3600Mhz Cas16
- PSU : Seasonic Prime PX-850, 850W Plus Platinum 




 

ace50k, this is, of course, the exact same PSU that I'm using, and the one that SCAN profess to have worked absolutely fine. I just don't understand why you should need to disable anything if you're running at stock.


@ace50kwrote:

It turns out the PSU was not able to handle the system properly as I swapped it out for an old 2008 (!) Corsair HX1000 I had in a spare PC and the system has been rock-solid since.  I've disabled PBO/Curve Optimizer for now and am running Cinebench's stability test as that used to cause WHEA errors/BSODs before.




 

kneel420's comment had me wondering if house electricals has anything to do with this?

Becuase one thing that I know is that (I'm aware how bad this is) the grounding in the office I'm plugged into isn't amazing. I do feel for you, though. I'm X570, too, a lot of responses in thread have listed full (as poss) specs, in the long run it could really help, but it won't mean much to most of us.


@kneel420wrote:

I'd be surprised if it was anyones GPU as people with both AMD and Nvidia GPUs had this issue.  Also someone fixed it with a new PSU?   I am at a loss with these reboots and its pretty frustrating.  Going to try the curve thing, then will try to replace the 3080 TI (good luck, they prob wont find one) then will try to replace the CPU.  I honestly thought it was the gigabyute X570 Aorus Master motherboard but peoiple in threads saying they had it on multiple boards including MSI and Asus.  Others claimed they RMA'd their 5950X's with AMD 3x and all of them still had the issue.   Not sure which to believe anymore!  Meanwhiile my PC is liable to reboot at anytime and these sys crashes are NOT graceful.




 

ZiN, I would pontificate around whether it's to do with how the motherboards handle the power, here. All the BIOS updates in the world won't mean a thing if the actual problem is specific to the manufacturer's processes (and that is happening cross company). However it is unlikely, because we have seen this reported across various AMD platforms, on multiple manufacturers, and the only thing that has been gradually improving things are BIOS updates.

Might be time for us to start a couple of threads at ASUS/MSI/etc (and maybe memory), just to ensure that there's parity there, to ensure that they even *know* this is a thing.

I don't believe for a second that we're just a few isolated cases on the internet, I believe most people will just either ignore it or never trigger whatever is doing it.


@Cmdr-ZiNwrote:

My issue has returned.

No issues for a couple of months after replacing my 5800X with a new one.

However out of the blue it crashed again on the 8th and again on the 10th.

I'm a bit tired of this now, I've never had this much issue with a platform ever before.

PSU was fine in my old system that drew more power, however maybe the new system needs something more sensitive. However there's no recommendations or specifications saying so.

The issues is most likely something to do with the MOBO, CPU, Gen4 NVME or 3600 MHz ram which I got at the same time.

The ram is no longer on the QVL list so maybe that's it but I would not of thought that would cause power cuts. The PC power light and RGB remain on but everything else power cuts which is strange.

Strange how this issues seems to follow infinity fabric. It's just so hard to narrow down.

Maybe the issue is just super random and rare but affecting all infinity fabric systems to some degree.



koguma/ZiN: I think raising it with the mem folks is a good (as above) idea, however, ensure that in your core, first, post, you reference this (long as heck) thread, so the context is clear. Also, as previously mentioned a lot ... including specs, triggers, and references to others with the same incident seen can increase the likelihood of it being taken more seriously.

Also, I spoke to SCAN about QVL specifically, and they stated that the RAM chosen for my setup (in a previous post) was QVL, however I wasn't totally sure.

One thing that I'm definitely thinking could be a wild shot in the dark is alerting one of the tech players on YouTube about this. They likely won't respond, but in the rare chance that they do, perhaps LTT, HWU, Gamers Nexus, and the grey haired bloke, might be able to 'raise awareness' ... or at least get us a specially coloured ribbon to wear.


@kogumawrote:

I've been ok for nearing the past two weeks by setting my Ram speed down to 3333 via the "Try It" MSI bios option.   Most likely it's an issue of really tight timings and/or specific timings.   My ram's XMP was specifically for Intel (I bought Intel certified ram, not AMD) so XMP always fails.  That means I can't run my ram at the specified 3600 speed.  I should probably reach out to G.Skill and ask them about this...



@Cmdr-ZiNwrote:

I'd love to hear what they say. I checked the QVL before I bought but maybe I messed up and bought the wrong stick or maybe they removed them from the list.

Still you'd think it wouldn't be that difficult, a quality stick of ram should be on the QVL list.


 

eliotcole
Adept II

Re: Ryzen 5900x: System constantly crashing/restarting WHEA-Logger ID 18 and critical error Kernel-P

Jump to solution

I would say, though, that the NVidia user experience is many multiples worse than this Radeon software that I'm now seeing on this card. Added to which there's quality of life stuff that I've always had issues with Nvidia (especially on sound) that I just really needed to switch. I'm happier now (not compensating ), but I don't want this to be a system breaker, y'know?


@Soulsa wrote:

It sucks, I'm over the moon AMD are doing fairly well BUT whenever I have had Intel they always just worked....NVIDIA also.... My 5700xt has had DRIVER problems for over 6 months, AMD KNOW also and we bombard them about it all.

 

I can't help but think it's always the driver team letting down the products...




I wonder if it's then down to differences in the CPU manufacturing. There could be very subtle differences, and these BIOS updates are tuning the system for the ones that they know are getting a pernickety. They're just addressing each manufacturing line one by one, hoping to eventually cover everone?


@Cmdr-ZiN wrote:

@Soulsa It's all been covered many times. It works for some, not everyone. However it's a work around not a fix.

Ryzen with infinity cache just can be unstable with various combination of components and narrowing them down is difficult because for most of the time it's stable.

I heard a story of 2 friends they built identical systems, one had the issue the other didn't. They swap CPU's with each other to see if the issue followed the CPU. However after they swapped CPU's, neither of them had the issue again.

I also heard Ryzen 3800X's had the issue and it was eventually fixed for most by a BIOS update.

For others it could be fixed by a CPU, GPU, PSU, MOBO or RAM swap, I've even heard of a sound card causing it.

I've thought I've had it fixed multiple times only to have it reappear months later. I still want to get to the bottom of it, however in all the systems I've ever owned, both AMD and Intel, I've never had an issue like this.




Mate, I would recommend highly, if you can, listing your PC specs, and if possible, whether you have found a way to capably replicate the incident. You don't have to, of course, but if this ever *does* become a thing, your input could really help.


@Soulsa wrote:

I was getting the same errors and had always thought it was GPU problems until I came across a similar story to mine from black screen restart while watching Youtube....

 

I don't know if you have tried these settings as this is a now 85 page forum post but could be low load restart issue.

As quoted.....
1. Disable C-stage power
2. Set PCIE to 3.0
3. Power supply set to Typical Idle.
SOC and Vcore all auto, XMP enable at default for you ram, CPB and PBO enable.

I don't have PBO on though myself, the power ramp for the minor speed rewarded isn't worth the hassle imo