CS2 Causing WHEA 18 Cache Hierarchy Error & Instant Reboots on Ryzen 5600X

[HMHEM]

I'm experiencing a very unusual issue that only occurs in CS2 and would appreciate any help or similar experiences.

System Specs:

CPU: Ryzen 5 5600X
Motherboard: MSI MAG B550 Tomahawk MAX WiFi
GPU: RX 6650 XT
RAM: 32GB (2×16GB) DDR4 3200
PSU: Corsair RM850x Shift 850W
UPS: Artis 2000VA
BIOS: Latest (v2.90)
Windows 11

Issue:
CS2 occasionally causes an instant black screen and automatic reboot. After reboot, Event Viewer shows:

WHEA Logger 18

Reported by component: Processor Core
Error Source: Machine Check Exception
Error Type: Cache Hierarchy Error
APIC IDs seen: 0, 1, 4, 5, 6, 7, 8, 9, 10

Troubleshooting already done:

BIOS updated
RAM moved to A2/B2
PBO disabled
CPB disabled
ReBAR disabled
CS2 reinstalled
FxSound uninstalled
SFC / DISM repaired Windows
Windows Memory Diagnostic passed
OCCT CPU test passed
OCCT Memory test passed
GPU stress test passed
CPU temperatures stay below 60°C while gaming

Interesting observations:

Forza Horizon 5 runs perfectly with no crashes or WHEA errors.
Offline CS2 is generally stable.
Crashes seem more likely during online CS2 (profile/inventory loading).
Lower resolution (1366×768) appears more stable and I've completed Premier matches without issues.

Has anyone with a Ryzen 5000 CPU or RX 6650 XT experienced similar WHEA 18 Cache Hierarchy errors only in CS2?

[Hyper Yoshino]

i had the same error with an intel 13th gen chip when my undervolt wasnt stable enough for cs but didnt crash with anything else, check that out maybe

[HMHEM]

Originally posted by j:

i had the same error with an intel 13th gen chip when my undervolt wasnt stable enough for cs but didnt crash with anything else, check that out maybe

Thanks for the suggestion. I checked my BIOS and everything is running at stock/default settings. I am not undervolting, overclocking, or using Curve Optimizer.

I've already tested with:

Latest BIOS
PBO disabled
CPB disabled
ReBAR disabled
Fresh CS2 install and new 730 folder
Windows Memory Diagnostic passed
OCCT CPU and Memory tests passed
GPU stress test passed
CPU temperatures remain below 60°C

The issue appears to be specific to CS2, especially during online/profile loading.

Event Viewer consistently reports:
WHEA Logger 18 → Machine Check Exception → Cache Hierarchy Error, with different APIC IDs (0,1,3,4,5,6,7,8,9,10).

If anyone with a Ryzen 5000 CPU has experienced something similar in CS2, I'd appreciate any suggestions.

[Hyper Yoshino]

Originally posted by Hunter Meets Hunter Eats Hunter:

Originally posted by j:

i had the same error with an intel 13th gen chip when my undervolt wasnt stable enough for cs but didnt crash with anything else, check that out maybe

Thanks for the suggestion. I checked my BIOS and everything is running at stock/default settings. I am not undervolting, overclocking, or using Curve Optimizer.

I've already tested with:

Latest BIOS
PBO disabled
CPB disabled
ReBAR disabled
Fresh CS2 install and new 730 folder
Windows Memory Diagnostic passed
OCCT CPU and Memory tests passed
GPU stress test passed
CPU temperatures remain below 60°C

The issue appears to be specific to CS2, especially during online/profile loading.

Event Viewer consistently reports:
WHEA Logger 18 → Machine Check Exception → Cache Hierarchy Error, with different APIC IDs (0,1,3,4,5,6,7,8,9,10).

If anyone with a Ryzen 5000 CPU has experienced something similar in CS2, I'd appreciate any suggestions.

Ah damn, I do have a Ryzen 5000 cpu but haven't experienced such an error (and tbh I run that pc on Linux most of the time), I hope other posters can be of help for you, good luck

I'd say this error lowkey sounds like a defect CPU to me, but it only happens with CS so that's weird

Can you like stress test with a tool like OCCT (it's on steam for free) and see if your PC crashes?

In OCCT try running

Memory
CPU+RAM
Linpack
And CPU and see if you get any errors

[HMHEM]

Originally posted by j:

Originally posted by Hunter Meets Hunter Eats Hunter:

Thanks for the suggestion. I checked my BIOS and everything is running at stock/default settings. I am not undervolting, overclocking, or using Curve Optimizer.

I've already tested with:

Latest BIOS
PBO disabled
CPB disabled
ReBAR disabled
Fresh CS2 install and new 730 folder
Windows Memory Diagnostic passed
OCCT CPU and Memory tests passed
GPU stress test passed
CPU temperatures remain below 60°C

The issue appears to be specific to CS2, especially during online/profile loading.

Event Viewer consistently reports:
WHEA Logger 18 → Machine Check Exception → Cache Hierarchy Error, with different APIC IDs (0,1,3,4,5,6,7,8,9,10).

If anyone with a Ryzen 5000 CPU has experienced something similar in CS2, I'd appreciate any suggestions.

Ah damn, I do have a Ryzen 5000 cpu but haven't experienced such an error (and tbh I run that pc on Linux most of the time), I hope other posters can be of help for you, good luck

I'd say this error lowkey sounds like a defect CPU to me, but it only happens with CS so that's weird

Can you like stress test with a tool like OCCT (it's on steam for free) and see if your PC crashes?

In OCCT try running

Memory
CPU+RAM
Linpack
And CPU and see if you get any errors

did already occt test for 1hr in memory cpu+ram, CPU Gpu everything fine 0 error no reboot

[Hyper Yoshino]

Originally posted by Hunter Meets Hunter Eats Hunter:

Originally posted by j:

Ah damn, I do have a Ryzen 5000 cpu but haven't experienced such an error (and tbh I run that pc on Linux most of the time), I hope other posters can be of help for you, good luck

I'd say this error lowkey sounds like a defect CPU to me, but it only happens with CS so that's weird

Can you like stress test with a tool like OCCT (it's on steam for free) and see if your PC crashes?

In OCCT try running

Memory
CPU+RAM
Linpack
And CPU and see if you get any errors

did already occt test for 1hr in memory cpu+ram, CPU Gpu everything fine 0 error no reboot

This lowk just sounds like a CS problem to me especially if it only happens in counter strike, not sure what you could do atp

[Mr. Rubber Ducky]

your CPU silicon is degraded if you have everything on stock and it still happens, or you're applying too aggressive of a negative offset via curve optimizer which reduces voltage and your chip is more power hungry.

The fact you only see it on CS2 is because it's mostly light on CPU usage which allows it to boost higher on single-core which for you will cause instability you're seeing. Not issue of the game but your CPU needs more voltage on those cores listed in your WHEA report, do Curve Optimizer positive +2-5 on affected cores and try playing it.

If you want to be 100% sure then leave your PC idle for hours or do something simple like exporting a video in Premiere Pro on GPU, you will see a hard reboot happen after a bit.

previous commenter was wrong, uv will only make it more unstable in this case. CHE means majority of the time insufficient voltage on core X and I can hit this error myself if I make too aggressive of a PBO2 config. I have a R9 5950X nowadays so ye.

[HMHEM]

Originally posted by Mr. Rubber Ducky #pride:

your CPU silicon is degraded if you have everything on stock and it still happens, or you're applying too aggressive of a negative offset via curve optimizer which reduces voltage and your chip is more power hungry.

The fact you only see it on CS2 is because it's mostly light on CPU usage which allows it to boost higher on single-core which for you will cause instability you're seeing. Not issue of the game but your CPU needs more voltage on those cores listed in your WHEA report, do Curve Optimizer positive +2-5 on affected cores and try playing it.

If you want to be 100% sure then leave your PC idle for hours or do something simple like exporting a video in Premiere Pro on GPU, you will see a hard reboot happen after a bit.

previous commenter was wrong, uv will only make it more unstable in this case. CHE means majority of the time insufficient voltage on core X and I can hit this error myself if I make too aggressive of a PBO2 config. I have a R9 5950X nowadays so ye.

Thanks, I think you were onto something. After more testing, it does seem related to voltage/stability rather than temperatures or obvious hardware failure.

My CPU is running completely at stock settings—no undervolt, no overclock, and no Curve Optimizer. I also noticed that CS2, especially during online/profile loading, is much more likely to trigger WHEA 18 Cache Hierarchy Errors.

OCCT CPU/Memory tests and GPU stress tests all pass, which makes me think CS2's boost behavior may be exposing a marginal voltage or power stability issue on my Ryzen 5 5600X.

I'm still testing, but voltage/boost instability currently seems like the most likely explanation.

[HMHEM]

Claude Helped xD.. and it worked NO crash since 4 hour intense gaming

The fix is a registry entry:
Press Win+R → type regedit → navigate to:
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Session Manager\Memory Management
Right click → New → DWORD (32-bit) → name it FeatureSettingsOverride → value 3
Right click → New → DWORD (32-bit) → name it FeatureSettingsOverrideMask → value 3
Restart.

What Was Actually Happening This Whole Time
CPU was perfectly fine. The real cause was:

AMD fTPM bug + CS2 VAC kernel driver = fake WHEA errors
Every time I went online, VAC's kernel-level anti-cheat accessed the TPM repeatedly. AMD's firmware TPM on Ryzen 5000 has a known bug where it causes random CPU stalls that Windows misreports as Cache Hierarchy Errors — making it look like a CPU hardware fault when it wasn't.
The two registry DWORDs you added disable the interaction that was triggering it.


Hope this helps if facing same issue :)

[PateaRojos]

Same issue here, 10s-1min running it and reboots my pc.
CPU AMD Ryzen 7 5700X 8-Core Processor 3.40 GHz
RAM 32 GB 3200 MHz
GPU AMD Radeon RX 9060 XT (16 GB)
Motherboard: B450 TOMAHAWK MAX II (MS-7C02)
Windows 10

I try your method but it didnt work :(
Is there any different config to apply for my components or could it be a w10 issue?

[HMHEM]

Can you try two tests?

1. Direct Wall Test

* Shut down the PC.
* Bypass any UPS/inverter/stabilizer and plug the PC directly into the wall outlet.
* Launch CS2 and see if the reboot still happens.

2. Check Input Voltage

* If you have a UPS with LCD or a voltage meter, check the wall voltage while gaming.
* My local voltage sometimes drops below 200V, so I'm trying to rule out power quality issues.

Also, after the reboot, please check:

**Event Viewer → Windows Logs → System**

Look for:

* WHEA-Logger Event ID 18
* Cache Hierarchy Error
* APIC ID

If you find a WHEA 18 entry, please post the APIC ID and error details.

I'm trying to determine whether this is a Ryzen 5000 series voltage/cache stability issue, a motherboard issue, or a power-quality issue affecting multiple users.

[PateaRojos]

Originally posted by Hunter Meets Hunter Eats Hunter:

Can you try two tests?

1. Direct Wall Test

* Shut down the PC.
* Bypass any UPS/inverter/stabilizer and plug the PC directly into the wall outlet.
* Launch CS2 and see if the reboot still happens.

2. Check Input Voltage

* If you have a UPS with LCD or a voltage meter, check the wall voltage while gaming.
* My local voltage sometimes drops below 200V, so I'm trying to rule out power quality issues.

Also, after the reboot, please check:

**Event Viewer → Windows Logs → System**

Look for:

* WHEA-Logger Event ID 18
* Cache Hierarchy Error
* APIC ID

If you find a WHEA 18 entry, please post the APIC ID and error details.

I'm trying to determine whether this is a Ryzen 5000 series voltage/cache stability issue, a motherboard issue, or a power-quality issue affecting multiple users.

I opened to see if i could look for the error in the event viewer, but it refused to crash for 3 hours playing straight, i opened a case and randomly crashed.
Then I checked the event viewer and i find the WHEA-Logger error with the ID 18:

Unrecoverable hardware error.

Reported by component: Processor Core
Error Source: Machine Check Exception
Error Type: Cache Hierarchy Error
Processor APIC ID: 13

AND THE DETAILS:
Log Name: System
Source: Microsoft-Windows-WHEA-Logger
Date: 21/06/2026 22:35:43
Event ID: 18
Task Category: None
Level: Error
Keywords:
User: LOCAL SERVICE
Computer: DESKTOP
Description:
A fatal hardware error has occurred.

Reported by component: Processor Core
Error Source: Machine Check Exception
Error Type: Cache Hierarchy Error
Processor APIC ID: 13

Additional information is available in the details view of this event.

Event XML:

```xml
<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="2026-06-21T20:35:43.9303101Z" />
<EventRecordID>311435</EventRecordID>
<Correlation ActivityID="{ec959981-6f9d-46ff-a178-cf24781cfd8b}" />
<Execution ProcessID="4312" ThreadID="4984" />
<Channel>System</Channel>
<Computer>DESKTOP-P45KITR</Computer>
<Security UserID="S-1-5-19" />
</System>
<EventData>
<Data Name="ErrorSource">3</Data>
<Data Name="ApicId">13</Data>
<Data Name="MCABank">5</Data>
<Data Name="MciStat">0xbea0000001000108</Data>
<Data Name="MciAddr">0x7ff92145ce9b</Data>
<Data Name="MciMisc">0xd01a0ffe00000000</Data>
<Data Name="ErrorType">9</Data>
<Data Name="TransactionType">2</Data>
<Data Name="Participation">256</Data>
<Data Name="RequestType">0</Data>
<Data Name="MemorIO">256</Data>
<Data Name="MemHierarchyLvl">0</Data>
<Data Name="Timeout">256</Data>
<Data Name="OperationType">256</Data>
<Data Name="Channel">256</Data>
<Data Name="Length">936</Data>
<Data Name="RawData">[RawData omitted here for readability - use the original XML if required]</Data>
</EventData>
</Event>
```

[HMHEM]

One thing I'd check is whether the Processor APIC ID stays the same or changes between crashes.

If the APIC ID is always the same core, it can point to a specific weak CPU core.

If the APIC ID changes between crashes (for example 0, 3, 8, 13, etc.), then it's less likely to be a single bad core and more likely a broader CPU stability or voltage/power issue.

In my case the APIC ID changes between crashes. I also discovered my area voltage fluctuates heavily (184–205V most of the time), and gaming on inverter power is noticeably more stable than direct mains power.

I'd suggest checking:

1.House voltage during gaming
2.Any bios setting you have made like, undervolt, Curve Optimizer.. reset to auto and make sure to update latest bios & pbo disabled in bios.
3.Whether the APIC ID changes across multiple WHEA 18 events

If the APIC ID keeps changing, I'd investigate overall CPU stability and power quality before assuming a specific CPU core is defective.

[PateaRojos]

i have the latest bios version, and checking the APIC ID i see it changes like 2, 10, 14, 3...

[HMHEM]

The fact that your APIC ID changes between crashes (2, 10, 14, 3, etc.). Since it keeps changing, I'd look at overall CPU stability, power delivery, or power quality instead. Chour your house voltage .. or psu. run OCCT test for 15 min cpu in extreme test to verify cpu is not the problem .... if your CPU and PSU is fine and Stable voltage you are getting then just mail to [e-mail] hope they help in this issue

[HMHEM]

one last trick you can try go in bios and disable Global C state.. i see amd new driver updates causing the issue