PCI express errors/crashes kernel erro41

seanuk · 8 Mar 2024 at 00:08

I keep experiencing crashes with the following hardware:

cpu: intel i9 14900k
ram: gskill 7200 tried running xmp and xmp off still crashing
gpu: ASU’s 4090

I keep getting intermittent crashes and upon checking the error log it says pci express kernel error 41 can anyone direct me in the right direction? Also ran numerous benchmarks with xmp on and off and noticed that if I lower my p cores in xtu from x57 to x54 it tends to run the tests without crashing could it possibly be a bad cpu causing me issues?

Tetras · 8 Mar 2024 at 00:33

seanuk said:
I keep getting intermittent crashes

What do you mean by "crash", can you elaborate? E.g. BSOD, CTD, black screen?

seanuk · 8 Mar 2024 at 07:30

Tetras said:
What do you mean by "crash", can you elaborate? E.g. BSOD, CTD, black screen?

Black screen buddy and upon checking event viewer it says kernel error 41, read numerous posts online where people have had issues with these cpu’s and had to lower the performance core ratio from x57 to x54 which I think is silly why buy a chip that you have to downclock? Tried cinebench r15 and it crashes sometimes at stock ratios but when down clocked to x54 it works? Any ideas anyone or does it look like I have a faulty cpu

mickyflinn · 8 Mar 2024 at 07:37

Bios up to date ?

seanuk · 8 Mar 2024 at 07:41

mickyflinn said:
Bios up to date ?

Yes latest buddy as I say can work for hours then next minute black screen

mickyflinn · 8 Mar 2024 at 07:53

seanuk said:
Yes latest buddy as I say can work for hours then next minute black screen

I would run memtest to rule the memory out .

seanuk · 8 Mar 2024 at 08:30

I’ve done that buddy that’s why I’m stumped exhausted most avenues I’m going to set the stock 253 power limit see if that changes anything incaste the cpu is pulling to much power causing the 12v to be unstable

mickyflinn · 8 Mar 2024 at 11:05

seanuk said:
I’ve done that buddy that’s why I’m stumped exhausted most avenues I’m going to set the stock 253 power limit see if that changes anything incaste the cpu is pulling to much power causing the 12v to be unstable

What psu you running?

seanuk · 8 Mar 2024 at 11:26

1000watt gold rated ASU’s been fine for months only just started to happen done fresh install also. I’ve applied 253 power limits now to see how that goes

mickyflinn · 8 Mar 2024 at 11:32

seanuk said:
1000watt gold rated ASU’s been fine for months only just started to happen done fresh install also. I’ve applied 253 power limits now to see how that goes

Yes been reading up on this myself which motherboard you got ?

wookiee87 · 8 Mar 2024 at 11:33

are you using a gpu sag bracket, the black screen and kernel 41 error is what was going on with my old setup, i ran a 4090 and now a 7900xtx.

turned out gpu sag had damaged my pcie socket on my old motherboard, when i changed it for a new one the black screens and crashes stopped, i also bought a sag bracket for my 7900xtx to avoid the costly mistake in the future

seanuk · 8 Mar 2024 at 12:57

mickyflinn said:
Yes been reading up on this myself which motherboard you got ?

Asus apex encore buddy

therealdippy · 8 Mar 2024 at 13:28

Error 41 is after the event. All it tells you is what you already know - as a result of the crash the PC did not shut down "gracefully". It is the errors before shutdown which count.
Instead of "try this", "try that", how about diagnosing the problem? Make sure that Windows is configured to produce a MEMORY.DMP file when it crashes. Then get WinDbg from the SDK and diagnose the dump. There is a good chance there will be something in the result which gives a strong indication as to the cause.

seanuk · 8 Mar 2024 at 13:47

Already on the latest bios buddy

seanuk · 8 Mar 2024 at 13:49

therealdippy said:
Error 41 is after the event. All it tells you is what you already know - as a result of the crash the PC did not shut down "gracefully". It is the errors before shutdown which count.
Instead of "try this", "try that", how about diagnosing the problem? Make sure that Windows is configured to produce a MEMORY.DMP file when it crashes. Then get WinDbg from the SDK and diagnose the dump. There is a good chance there will be something in the result which gives a strong indication as to the cause.

How do I configure it produce a memory.dmp when it happens buddy?

mickyflinn · 8 Mar 2024 at 15:37

seanuk said:
How do I configure it produce a memory.dmp when it happens buddy?

Generate a kernel or complete crash dump - Windows Client

Learn how to generate a kernel or complete crash dump, and then use the output to troubleshoot several issues.

learn.microsoft.com

As @wookiee87 mentioned do you have anti sag bracket for the GPU as the strain on the pci socket can cause problems.

seanuk · 8 Mar 2024 at 15:51

mickyflinn said:
Generate a kernel or complete crash dump - Windows Client

Learn how to generate a kernel or complete crash dump, and then use the output to troubleshoot several issues.

learn.microsoft.com

As @wookiee87 mentioned do you have anti sag bracket for the GPU as the strain on the pci socket can cause problems.

Yes I have an anti sag bracket on card is sitting nice no sag on it. Since I’ve enabled a power limit of 253 on the cpu it’s been ok upto now

wookiee87 · 8 Mar 2024 at 16:46

Been looking over the specs of the 14900k, it's effective boost clock is 5.6ghz, then boost 3.0 will let the chip go to 5.8ghz and finally thermal velocity boost will let 1/2 cores go to 6.0ghz.

I would say your chip is fine, problem is intel and AMD don't validate every chip to there max speed they can reach, as you say if many other owners have had to reduce the ratio to 54 to maintain stability, you effectively lost the silicone lottery.

I'm the same my 7950x will boost into the mid 5.6ghz region on 1/2 cores but never at its rated 5.7ghz boost it should do, like you I lost the silicone lottery as well as many many others have.

As long as your chip is inline or better than base frequency there isn't much you can do about it unless the CPU is faulty.

seanuk · 8 Mar 2024 at 17:41

wookiee87 said:
Been looking over the specs of the 14900k, it's effective boost clock is 5.6ghz, then boost 3.0 will let the chip go to 5.8ghz and finally thermal velocity boost will let 1/2 cores go to 6.0ghz.

I would say your chip is fine, problem is intel and AMD don't validate every chip to there max speed they can reach, as you say if many other owners have had to reduce the ratio to 54 to maintain stability, you effectively lost the silicone lottery.

I'm the same my 7950x will boost into the mid 5.6ghz region on 1/2 cores but never at its rated 5.7ghz boost it should do, like you I lost the silicone lottery as well as many many others have.

As long as your chip is inline or better than base frequency there isn't much you can do about it unless the CPU is faulty.

I’ve set the recommended power limit of 253watt and haven’t had any crashes yet so makes me think was it pulling to much power therefore causing it to get pci errors?

wookiee87 · 8 Mar 2024 at 17:53

You could be on to something if the CPU wasn't stable at oob power settings it could have affected signal integrity between it and a dedicated GPU, though saying that and the behaviour you were getting is very similar to what I was getting, during your crashes before did you ever get in game messages like "pcie device removed, this happens if the GPU is disconnected" I had that plus loads of error 41's in event viewer on my old setup.