Build Advice Fresh Desktop Build Consistently Crashes; Can't Figure Out Why

[LogixTheDev]

Hi all.

I recently built a desktop PC from a part list that was put together by a friend. It's the first time I've built a computer from scratch, but not my first time tinkering around with one.
Anyway, I believe the whole thing to be put together correctly; it's not that complicated after all. All the parts are brand new.

It runs just fine most of the time, but it consistently crashes whenever I try to run literally any game, or various programs.

• When playing any game, it will be guaranteed to crash within a few moments to about 20 minutes or so. Some games run for longer, but only on rare occasions
• When using the browser to stream a video, sometimes the page will just randomly stop working (if this happens again I'll update the post with the error it displays, because I can't remember off the top of my head
• When programming, my IDE will just suddenly tell me there was an error (not what the error was, though) and that it needs to restart
• During some game crashes, the game will just close. Other times the audio glitches and the whole PC crashes, either to black followed by a restart, or occasionally a BSOD
• I've also had Steam restart itself during gameplay once, and it just kept trying to open and then closing itself over and over again with the game still running (apparently just fine) until I managed to painstakingly kill it with Task Manager (timing was everything!)

Frustratingly, almost every time I get a BSOD it's a different error, which is why I'm posting this.
I've tried everything to diagnose / fix the problem, and everyone I've spoken to about it or shown the computer to is as baffled as to what could be wrong with it as I am.

Here's a list of some of the BSOD error codes I've had:

• SYSTEM_THREAD_EXCEPTION_NOT_HANDLED (NetAdapterCx.sys)
• SPECIAL_POOL_DETECTED_MEMORY_CORRUPTION
• PAGE_FAULT_IN_NONPAGED_AREA
• MEMORY_MANAGEMENT
• DRIVER_IRQL_NOT_LESS_OR_EQUAL (dxgmms2.sys; ntoskrnl.exe)
• REFERENCE_BY_POINTER
• KERNEL_SECURITY_CHECK_FAILURE
• DRIVER_VERIFIER_DETECTED_VIOLATION
• ATTEMPTED_TO_WRITE_TO_READONLY_MEMORY

Some of these codes were presented before I bought new RAM, thinking that was the issue (and because the RAM I was using was placeholder until I could afford the more expensive sticks I wanted).

Here are some of the (many) things I've tried:

• Re-Installing Windows from an OEM CD and, in separate instances, using the downloaded ISO from the Microsoft website
  • I've tried installing Windows on 3 separate drives. Now running it on a brand new SSD slotted directly into the motherboard
• Running filesystem scans (/sfc scannow and the DISM one)
• Ensuring Windows is up-to-date
• Using Driver Booster Pro to ensure all drivers are up-to-date
• Updating the BIOS
• Resetting the BIOS to default settings and ensuring overclocking is disabled
• Trying different memory profiles (A-XMP, EXPO) to see if the memory was running at the wrong speed (I don't know much about this, though)
• Re-building the entire computer from the ground up and checking all cables and components for damage and correct placement
• Seating the RAM in different slots (it's back to the recommended slots)
• Running memtest86 (0 errors found)
• Stress-testing the CPU and GPU (no problems)
• Monitoring power usage and temperatures
• ...And many more!

The computer doesn't overheat (to the best of my knowledge); graphs show decent temperatures and the air coming out of it is never particularly hot. This shouldn't be a concern as the processor has a liquid cooling block attached with thermal paste applied.

The computer has sufficient memory (32 GB) and I've tried 2 sets of brand-new RAM sticks, they seem absolutely fine.

I suspect that it could be driver issues or a faulty component (REALLY hoping it's not the latter as I've spent enough on this cursed thing!).

I'm looking for assistance from someone with a little patience, so that I can get the machine up and running and return to normality. I'm missing out on several expansions that have just released on a few games, and unable to work on my programming projects, and already have limited time due to work.

Here are the specs, in case they're relevant:

• Motherboard: MSI B650 Gaming Plus WiFi AX AM5
• PSU: Corsair RM850e 850W 80+ Gold Certified Fully Modular ATX Power Supply
• GPU: MSI Ventus 3X OC GeForce RTX 4060 Ti 8 GB
• CPU: AMD Ryzen 5 7600X 6-Core AM5
• RAM: G.Skill Trident Z5 RGB DDR5 (2 x 16 GB) 6000
• Cooling: Corsair iCUE H150i Elite Capellix XT 65.57 CFM Liquid CPU Cooler

The PSU should be more than sufficient, according to PC Part Picker, but this is another thing I'm aware could be the problem.

I'm running Windows 11 23H2 (Build 22631.3810), and for graphics drivers, I typically use NVidia GeForce Experience to ensure I always have the latest game-ready drivers (especially important now that I've finally made the leap up to RTX).

If any more information is needed, ask away.

Thank you in advance!

[EDIT]
I managed to play Destiny 2 for all of about 3 minutes before it crashed, which was enough for me to glean some information. First of all, the GPU was making an ominous electrical buzzing sound while in use, which seemed to react to the game (the pitch or volume of the bussing shifts when looking around/switching weapon). During gameplay, Discord closed and re-opened itself. The game crashed about a minute later, wth the following log (cropped to protect some data which should not be shared):

stack:
00007FF7A92D0F19 <unknown>
00007FF7A8B65750 <unknown>
00007FF7A983E090 <unknown>
00007FF7A9D54FCD <unknown>
00007FF7A905F7A2 <unknown>
00007FF7A983E161 <unknown>
00007FF7A9600923 <unknown>
00007FF7A9600BB6 <unknown>
00007FF9FC58257D <unknown>
00007FF9FE88AF28 <unknown>
session_id:
Fri_07052024_195638_005220@x64_steam@76561198274457563@BD_CB_E6_D0_F1_58_7882
version:
v500_live.main tiger_final release pc_x64 226472.24.06.26.1731.v500_live.main 24.06.26 1731
thread information:
thread_name: unknown, thread_id: 00004d48
halt:
### RUNTIME ERROR: EXCEPTION_ACCESS_VIOLATION at 00007FF7A92D5124
halt information:
(<unknown>)
crash: tried to read address FFFFFFFFFFFFFFFF
fatal error detected, current status was:
init: init.txt
version: v500_live.main tiger_final release pc_x64 226472.24.06.26.1731.v500_live.main 24.06.26 1731

It seems it was trying to read a completely invalid memory address. In other crashes I've had similar execution violations and asset streaming failures. This would SEEM to implicate something to do with memory, however I'm fairly sure at this point that the memory isn't the issue. Unless this is to do with VRAM. However, one other thing I forgot to mention is that I tried running the game with the GPU removed, using integrated graphics. It was PAINFULLY slow, of course, but the crashes still occurred eventually. So, the GPU shouldn't be the problem either. I might have to try that one again with a different game just to verify, but for now it's not on the list of suspects.

 

[Ralston18]

This motherboard?

https://download.msi.com/archive/mnu_exe/mb/B650GAMINGPLUSWIFI.pdf

(Check to ensure that I found the applicable manual.)

DIMM Slots

When the build was assembled was the first physicallly installed RAM placed in DIMMA2 per physically numbered Page 28?

That is important.

 

[LogixTheDev]

This motherboard?

https://download.msi.com/archive/mnu_exe/mb/B650GAMINGPLUSWIFI.pdf

(Check to ensure that I found the applicable manual.)

DIMM Slots

When the build was assembled was the first physicallly installed RAM placed in DIMMA2 per physically numbered Page 28?

That is important.

Yes, that's the one. When I first assembled it, I had placed one of the two RAM sticks in slot DIMMB1 (I didn't yet know positioning was important, or what the markings on the motherboard meant), but the computer told me to move it upon boot, which I promptly did:

 

[Ralston18]

Try a "do over".

Power off, unplug.

First remove both of the RAM modules.

Next clear CMOS per the User Manual - JBAT1, Page 41.

Then install RAM into DIMMA2 first, then DIMMB2.

Reboot.

Reconfigure only as necessary.

Immediate objective being to achieve a stable system.

= = = =

Intesting that the boot process flagged the non-optimized configuration. However, the solution may have been a bit oversimplified. Just as a general thought....

 

[LogixTheDev]

I have followed your suggestion. It looks like it went relatively smoothly. Some notes:

Upon boot, the EZ Debug LEDs lit up:
CPU: Red
DRAM: Amber
VGA: (Off)

After a few moments they changed to:
CPU: (Off)
DRAM: Amber
VGA: White

Then the system seemed to restart and booted up. It told me that the settings had been cleared and that I could enter the BIOS or load default settings. Before I had the chance to make a choice, it continued to boot up to Windows. My guess is it went with default settings.

As for the boot process flagging the configuration, I have never seen anything like it myself either. But this is modern MSI hardware. It even seems to run the BIOS screen on the GPU (!) ...All BIOS' I've used in the past have ignored the GPU and ran off integrated graphics until I boot into Windows and the drivers get loaded, but the GPU is the only thing plugged into the monitor and I can see the BIOS screen.

Anyway, what would be the next step you'd recommend?

 

[Ralston18]

Regarding "next step" - I recommend simply waiting through a few boot-ups and updates.

May take a bit of time for Windows etc. to get all cleaned up.

Continue watching error codes, etc.. Keep notes.

The objective being to discover some specific pattern of problems or other errors that may arise.

Watch carefully and monitor as much as you can.

And, as always, ensure that all important data is continually backed up somewhere away from the system in question.

 

[LogixTheDev]

Thanks. The computer's still behaving the same as before, and just blue-screened with a new error I haven't seen before: UNEXPECTED_STORE_EXCEPTION. This was after I tried to run Destiny again and it crashed after around 5 minutes, but this time the game had simply frozen. The music was still playing, the game had just frozen.

Thanks to this new stop code, I've found a post on Microsoft Answers* which looks like the OP was suffering similar behaviour, however, as is always the case with Msft Answers it's not helpful in the slightest. Just a whole lot of grasping at straws. In any case, the OP said their issue was a faulty RAM module. I'm sceptical that that's the problem I'm facing, as I've replaced the RAM with brand new ones less than a week ago, and the likelihood that they're faulty too is improbable. These sticks were also much better-packaged than the previous ones, with ESD packaging and soft padding to protect against damage.

Could it be the RAM ports on the motherboard itself? The motherboard does bend a little when pressing the RAM sticks in. I try to support it as much as possible to prevent it from deforming as much as possible, but I still have to press firmly to get the RAM sticks to actually insert far enough in for the clips to engage.

Other things I've noticed:

The PSU seems to be suffering from a little coil whine. It's not too bad, but it changes pitch when doing certain things such as scrolling on a webpage. I suppose this is linked with GPU power draw?

The GPU has started making a slight buzzing noise which gets quite a bit louder when running a game, and also seems to respond to how much rendering is happening on the screen -- it gets quieter if I open a static game menu, for example. It changes pitch as I hover over different characters on the character select screen.

Before most game crashes, it seems Windows Event Viewer is logging a warning from NVIDIA OpenGL Driver, stating:

DrvSetContext failed functionality indeterminant
(pid=14912 cncmd.exe 64bit)

(The process ID is different each time, of course). I feel like this is likely very relevant to the problem at hand.

I'm going to try a graphics driver update, as one has become available. I don't expect this will fix the issue, as I've already tried this a few times now to no effect.

*I couldn't link the post, for some reason the site flagged it as spam-like content . It's titled "Unexpected Store Exception Windows 11" by Matthew Wever, April 23 2023.