Question Hardware error during VS compilation procedure, should I RMA my CPU ?

[RedYggdrasil]

Hello everyone !

So I work on Unreal Engine and so must build the engine every now and then. The same task work flawlessly on my other (much older) computers.

But lastly I could not build the engine anymore on my main Workstation as Visual Studio constantly crashes with a lot of compiler internal errors :

1>C:\UE_B\UnrealEngine\Engine\Source\Runtime\Core\Public\Containers\Map.h(921): fatal error C1001: Internal compiler error.
1>(compiler file 'D:\a\_work\1\s\src\vctools\Compiler\Utc\src\p2\main.c', line 225)
1> To work around this problem, try simplifying or changing the program near the locations listed above.
1>If possible please provide a repro here: https://developercommunity.visualstudio.com
1>Please choose the Technical Support command on the Visual C++
1> Help menu, or open the Technical Support help file for more information

This continues until complete build failure or system freeze and the occasional BSOD.

In the Windows Event Viewer, each VS reported errors are preceded by a warning of corrected Hardware Failure :

A corrected hardware error has occurred.

Reported by component: Processor Core
Error Source: Corrected Machine Check
Error Type: Internal parity error
Processor APIC ID: 32

The details view of this entry contains further information.

I tried:

• Reinstalling the source code multiple times on different SSDs.
• Reinstalling different version of Visual Studio multiple times.
• I formatted my computer and reinstalled it.
• I modified my Windows power plan to not allow the CPU to go above 95% power usage.
• I ran a CHKdsk disk check and MemTest86, both without error.

The strange thing is that I fail to put my CPU at fault with other stress test, but the VS build procedure always make it go haywire. A shame considering than I bought it specifically for that task in January 2023.

My Workstation :

OS : Windows 11 Pro
Motherboard : ROG STRIX Z790-E GAMING WIFI (BIOS Updated yesterday as part of the troubleshooting process, no change)
CPU : Intel i9-13900K
RAM : G.Skill Ripjaws F5-6000J3040G32GX2-RS5K 64 Go 2 x 32 Go DDR5 6000 MHz
SSD : XPG GAMMIX S70 BLADE 2To x2 (tested on both)
HDD : FireCuda 8TB (not used)
GPU : NVIDIA GeForce RTX 3090

Also I saw (very few) other peoples having kind of similar problem a few months back, on of them at least was also using the Asus MB + i9 CPU combo.

Considering the Hardware error is always reported on "Processor APIC ID: 32" I am heavily tempted to think than the issue come from my CPU and that said core is cooked. But I don't know if what data I have collected so far is proof enough.

Thank you for having read me and I hope you may have insight as to what I should do next.
Not being able to work for now 3 days straight as I just released a product I cannot support is becoming increasingly unfun.

RedYggdrasil

 

[rgd1101]

what the cpu temp when it compiling?
brand and model of the psu?

 

[RedYggdrasil]

Hello and thanks for the reply !

Well now that's an intresting question isn't it ? quite telling than I didn't even do a check ... I think I'm kind of tired.

Regardless :

PSU :
Fox Spirit US-1000G 80PLUS Gold
By Seasonic Electronics for french reseller LDLC
https://www.ldlc.com/fiche/PB00409687.html

CPU Cooling :
Noctua NH-D15 chromax black

Thermal Paste :
Noctua NT-H2, applied in January 2023 (9 m/o)

During the build process Armoury Crate report :

Medium Performance mode, CPU cap at 95% Power usage, and PC Case windows open, Case out of it's enclosing space :

CPU Temperature ranging from 79° and 82° degree Celsus (179.6 Fahrenheit top)
CPU Package from 86° to 92° (197 Fahrenheit top)
5 crash error report during the build process (out 5.8 K Tasks)

High Performance mode, CPU Uncapped (but still no OCs), and case back into it's rat-hole under my desk :

CPU Temperature ranging from 86° and 88° degree Celsus (190 Fahrenheit top)
CPU Package from 90° to 99° (210 Fahrenheit)
Many crash error report during the build process, I canceled it before PC had a chance to crash

Is this within expected range or am I just a silly boy how forgot to change it's CPU's thermal Paste every 6 month or something ? Because in any way you're right, temperature seams to clearly play a big role there.

However There is still 5 error on medium perf test when I would have expected none (I don't have this happening on any other computer, and they have worst CPU fan).

RedYggdrasil.

 

[Bob.B]

Hello everyone !

So I work on Unreal Engine and so must build the engine every now and then. The same task work flawlessly on my other (much older) computers.

But lastly I could not build the engine anymore on my main Workstation as Visual Studio constantly crashes with a lot of compiler internal errors :

1>C:\UE_B\UnrealEngine\Engine\Source\Runtime\Core\Public\Containers\Map.h(921): fatal error C1001: Internal compiler error.
1>(compiler file 'D:\a\_work\1\s\src\vctools\Compiler\Utc\src\p2\main.c', line 225)
1> To work around this problem, try simplifying or changing the program near the locations listed above.
1>If possible please provide a repro here: https://developercommunity.visualstudio.com
1>Please choose the Technical Support command on the Visual C++
1> Help menu, or open the Technical Support help file for more information

This continues until complete build failure or system freeze and the occasional BSOD.

In the Windows Event Viewer, each VS reported errors are preceded by a warning of corrected Hardware Failure :

A corrected hardware error has occurred.

Reported by component: Processor Core
Error Source: Corrected Machine Check
Error Type: Internal parity error
Processor APIC ID: 32

The details view of this entry contains further information.

I tried:

• Reinstalling the source code multiple times on different SSDs.
• Reinstalling different version of Visual Studio multiple times.
• I formatted my computer and reinstalled it.
• I modified my Windows power plan to not allow the CPU to go above 95% power usage.
• I ran a CHKdsk disk check and MemTest86, both without error.

The strange thing is that I fail to put my CPU at fault with other stress test, but the VS build procedure always make it go haywire. A shame considering than I bought it specifically for that task in January 2023.

My Workstation :

OS : Windows 11 Pro
Motherboard : ROG STRIX Z790-E GAMING WIFI (BIOS Updated yesterday as part of the troubleshooting process, no change)
CPU : Intel i9-13900K
RAM : G.Skill Ripjaws F5-6000J3040G32GX2-RS5K 64 Go 2 x 32 Go DDR5 6000 MHz
SSD : XPG GAMMIX S70 BLADE 2To x2 (tested on both)
HDD : FireCuda 8TB (not used)
GPU : NVIDIA GeForce RTX 3090

Also I saw (very few) other peoples having kind of similar problem a few months back, on of them at least was also using the Asus MB + i9 CPU combo.

Considering the Hardware error is always reported on "Processor APIC ID: 32" I am heavily tempted to think than the issue come from my CPU and that said core is cooked. But I don't know if what data I have collected so far is proof enough.

Thank you for having read me and I hope you may have insight as to what I should do next.
Not being able to work for now 3 days straight as I just released a product I cannot support is becoming increasingly unfun.

RedYggdrasil

If your running the ram at 6000 bring it down to 5600 see if it makes a diff.

 

[RedYggdrasil]

If your running the ram at 6000 bring it down to 5600 see if it makes a diff.

Thanks for the reply ! and sorry for the delay on mine, I was traveling.
I've given up on XMP long ago, ram are running at stock DDR4 speed since pretty much day one (4800).