PC crashing since W11 upgrade

Phunky Phish · 30 Nov 2023 at 22:54

Sooooooo.....
I had a totally stable PC running windows 10, but I thought the time was right to upgrade to windows 11.
After upgradinng, my PC started crashing whilst gaming, which it never did before.
I have downloaded the latest bios, drivers, etc, but still crashes.
Did a full clean install of windows 11, still crashes.
Did a full clean install of windows 10, still crashes.
Disconnected a spare HDD, still crashes.
Took out 1 stick of RAM and tested, crashed. Tested other stick, crashed.
Changed the SSD, still crashed.

Here is the spec of the PC -
Mobo - MSI Tomahawk x570
CPU - Ryzen 5 5600X 4.6Ghz
RAM - 2x16GB Corsair Vengence LPX DDR4 3600C18
GPU - Gigabyte GeForce RTX 3080 10GB AORUS XTREME V2
SSD - WD 1TB SN750 M.2 2280
HDD - WD 2TB Black
PSU - Corsair RMx750

It seems to always crash after about 20-30 minutes of playing Fornite (normally when I am in the last 5!!).

I am at the point of having to start buying new PSU's, CPU's or motherboards, and I cant afford that right now, so any ideas would be really helpful.

Also, any ideas on what to use to stress individual parts would be good, so I could stress the CPU, try the memory again, etc, might help narrow this down.

Cheers for your help.
Mark

Tetras · 30 Nov 2023 at 23:50

Phunky Phish said:
It seems to always crash after about 20-30 minutes of playing Fornite (normally when I am in the last 5!!).

By "Crash" what exactly do you mean? Black screen? BSOD? CTD?

Phunky Phish · 1 Dec 2023 at 07:36

lol. Good point!

It normally crashes straight to a reboot, no BSOD.
The one time I got a BSOD I got WHEA_UNCORRECTABLE_ERROR, which is a hardware fault I believe (googled it)

Mcnumpty2323 · 1 Dec 2023 at 10:18

To test individual parts
Ram use memtest though need a USB flash drive
Cpu I use cinebench r23 single core
And multicore and run it for quite a while
This is more a real world test
For extreme test prime95
For gpu I use unigine,heaven,valley or superposition
Looped for as long as you like
For combined test
Both unigine and cinebench r23 at same time
For extreme gpu test furmark
Though don't remember if nvidia implemented some sort
Of power limit if their driver detects furmark running
For hard drive test
The manufacturers of the drive software
Should include a self test long and short

All of those are free
Personally I prefer real life test over extreme stress tests
Just my personal choice though
If everything passes could be psu
May just be coincidence it happened as you tried 11
Since clean install of 10
It should have been same as previously

Obviously check cpu and gpu temperature
While testing
Something like hwinfo, sidebar diagnostics etc
Will show both of them

Ch3m1c4L · 1 Dec 2023 at 11:53

If you're getting whea errors, it should be logged in event viewer, likely in system tab, might provide info as to which part is being reported faulty.

Phunky Phish · 1 Dec 2023 at 21:02

Ch3m1c4L said:
If you're getting whea errors, it should be logged in event viewer, likely in system tab, might provide info as to which part is being reported faulty.

I have a critical log for the Kernal-Power, but this might just be when it crashed?

Ch3m1c4L · 1 Dec 2023 at 21:13

If it just says the last system shutdown was unexpected, or it has booted after a bad shutdown, those two are normal for any crash.
There should be one with source of whea logger or whea crash reporter or whatever its called, it will normally show up above(more recently) in the log than those other kernal ones, after it's back on.

Can also check you haven't got instant auto restart on crash enabled via click on my computer, properties, advanced... Somewhere I've forgotten the name of

Did you upgrade/change bios or any cpu stuff like PBO or curve optimizer before the crashing started happening?

Phunky Phish · 1 Dec 2023 at 21:17

I have this against the WHEA log

- <Event xmlns="
http://schemas.microsoft.com/win/2004/08/events/event
">
- <System>
<Provider Name="Microsoft-Windows-WHEA-Logger" Guid="{c26c4f3c-3f66-4e99-8f8a-39405cfed220}" />
<EventID>18</EventID>
<Version>0</Version>
<Level>2</Level>
<Task>0</Task>
<Opcode>0</Opcode>
<Keywords>0x8000000000000000</Keywords>
<TimeCreated SystemTime="2023-11-28T21:47:45.6761836Z" />
<EventRecordID>1678</EventRecordID>
<Correlation ActivityID="{bc156a5e-aac4-4ba4-ac18-6428529d7373}" />
<Execution ProcessID="4292" ThreadID="4820" />
<Channel>System</Channel>
<Computer>DESKTOP-UKUHJTI</Computer>
<Security UserID="S-1-5-19" />
</System>
- <EventData>
<Data Name="ErrorSource">3</Data>
<Data Name="ApicId">11</Data>
<Data Name="MCABank">1</Data>
<Data Name="MciStat">0xbc800800060c0859</Data>
<Data Name="MciAddr">0x36643fc40</Data>
<Data Name="MciMisc">0xd01a0ffe00000000</Data>
<Data Name="ErrorType">10</Data>
<Data Name="TransactionType">256</Data>
<Data Name="Participation">0</Data>
<Data Name="RequestType">5</Data>
<Data Name="MemorIO">2</Data>
<Data Name="MemHierarchyLvl">1</Data>
<Data Name="Timeout">0</Data>
<Data Name="OperationType">256</Data>
<Data Name="Channel">256</Data>
<Data Name="Length">936</Data>
<Data Name="RawData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ata>
</EventData>
</Event>

Phunky Phish · 1 Dec 2023 at 21:27

Ch3m1c4L said:
If it just says the last system shutdown was unexpected, or it has booted after a bad shutdown, those two are normal for any crash.
There should be one with source of whea logger or whea crash reporter or whatever its called, it will normally show up above(more recently) in the log than those other kernal ones, after it's back on.

Can also check you haven't got instant auto restart on crash enabled via click on my computer, properties, advanced... Somewhere I've forgotten the name of

Did you upgrade/change bios or any cpu stuff like PBO or curve optimizer before the crashing started happening?

Hi. Thanks for your time.
All I did was upgrade to windows 11

Ch3m1c4L · 1 Dec 2023 at 21:56

I think it's likely due to a cpu core, at idle going into low power state and crashing because it feels like it, or doesn't like the voltage is being given, or not being given.

Things to try, in no particular order, just do the least invasive/easiest first and continue as needed:

You already updated bios.

Did you get latest chipset drivers from msi or amd? Confirm which has latest version posted and get that.

Disable any PBO/auto bios cpu boosting above stock.

If using curve optimiser to undervolt, stop, at least on the relevant core/s (apicid 11 so core 6 for this particular error I think), if not using curve optimiser, turn it on and bump that /all cores up by 5 or 10mv to see what happens. Basically bump the voltage curve up for the cpu or core 6 from whatever it currently is at.

Load line calibration could also be set at a higher level too to help if its vdroop related.

Turn off any/some c-states (low power states) for cpu in bios, and anything and everything in windows power management.

It's possible it's other hardware, could be anything that's about to die (that you havent disconnected), psu not delivering correct power or gpu not responding to a low level driver operation etc, but these are less likely as they *should* present differently but people have had this sort of error go away after random hardware swaps.

Phunky Phish · 1 Dec 2023 at 22:31

Well I ran the 30min Cinebench stability test. No issues. Temps were max 68c with average 66.1c, after a 10min then 30min test.

I’ll run a GPU test next.

Ch3m1c4L · 1 Dec 2023 at 22:36

If its a low power state problem, it won't show up during a stress test as then it's always going to be in a high power state

Phunky Phish · 1 Dec 2023 at 23:12

It only crashes when gaming. It wouldn’t go into a low power state during gaming would it?

30min gpu stress worked fine. Max gpu temp 60c

Ch3m1c4L · 1 Dec 2023 at 23:19

Yes, in some games, like fortnight, it will.
Cores constantly go up and down through power states and clock speeds all the time and except when under no load where they stay in lowest power state, or very heavy/full load where they sit at as close to max boost that temperatures allow in their highest power state (except for actual single core loads that might clock minorly higher).
But there are many inbetween nowadays, same as on the gpu, if one clock speed to voltage step is incorrect in your mobo settings or you've got one for a tiny bit weaker than it should be, it could show up like this.

Games are generally 1 to 8 threads, with 1 to 2 being heavy and the rest being lighter loads, and they bounce around all the cores/threads available as each one cycles up and down to optimise for highest boosts and lowest temps.

Phunky Phish · 1 Dec 2023 at 23:24

Ok. I’m gonna run the gpu and cpu stress test for 30mins. If all is fine, I’ll start checking the voltages.
The PSU has a 10 year warranty, so I could send that back to OCUK to check I guess.
Everything is run stock in the bios, no overclocking or undervolting, as far as I know!

Phunky Phish · 1 Dec 2023 at 23:34

Odd. Stressing both is giving me a max cpu temp of 71.3c, but average is lower than stressing only the cpu at 64.7c!

Phunky Phish · 1 Dec 2023 at 23:52

Stressing both passed for 30mins.
I’ll try checking the volts tomorrow.
Thanks for your help and suggestions ch3m1c4L

Phunky Phish · 3 Dec 2023 at 16:31

It looks like the cpu voltage is set to auto and 1.35v, so I have manually set it to 1.4v

Mcnumpty2323 · 3 Dec 2023 at 16:48

Only time I had the idle/low power state
crash issue
Was when I was messing around
With manual pbo offsets/curves
Though got a 5950x
So not the same cpu
But on default bios settings no issues

As someone already said
Could try turning off pbo in the bios
If its enabled

Though doesn't really explain the crashes
As presumably it was set the same to start with

Other thing to look at
Which windows power setting are you using?
Best performance should stop
Any downclocking

Phunky Phish · 3 Dec 2023 at 19:38

PBO is off.
I was on balanced but will go performance.
Since changing the cpu voltage to 1.4v it seems stable again.