Memtest Test 9, inconsistent errors.

[Lang]

Hello all,

I built my PC in Sep. of last year, and ever since the build I have had issues with the device. I have so many pictures of different blue screens in my camera roll it's not even funny. I will try to get everything gathered into one big message.

What I'm experiencing:

Often when playing games, specifically mildly resource-intensive games, my PC has a tendency to bluescreen or crash to the desktop after around 10-30 minutes of play. More often then not, it fully bluescreens providing a various bluescreen error code (IRQL_LESS_NOT_EQUAL or KERNEL_SECURITY_CHECK_FAILURE are the ones that come to mind from memory). Often if I quickly go to Event Viewer after the computer crashes I can find an "APP_CRASH" event or something along those lines. However, recently in event viewer I just find vague Windows Error Report events that site "Event Name: BlueScreen" (I will attach the memdump below).



What I have tried to fix:

Essentially everything... All versions of DISM and SFC scans, all CHKDSK scans, fully new RAM from a different manufacturer, a fully new GPU verified to work on other PCs, full new installs of Windows 11 & Windows 11 Pro, updated all drivers (GPU, CPU, Chipset, etc.), tried DDU to install new GPU drivers.



Device Specs:

- ASUS TUF 4080 GAMING

- Ryzen 7 7700x

- 850w PSU from Corsair

- 64GB DDR5 Corsair Vengance RAM,

- 240 ML AIO from Cooler Master

- B650 TOMAHAWK WIFI MOBO


Files for helping me :

https://langdongg.me/DIMMResults-20240407-144204.bmp
https://langdongg.me/MemTest86-20240407-100259.log
https://langdongg.me/mt86SysInfo-20240407-144301.txt
https://langdongg.me/MemTest86-Repor...07-100336.html

[David (PassMark)]

Link to .txt file and .log file are broken. That host doesn't seem to serve those file types?

From the HTML report:

• The CPU is running a little hot (max 90C), and this only while testing RAM. In gaming it might be even hotter.
• There was a couple of RAM errors. So 100% sure there is some type of hardware or config error. (So CHKDSK, new drivers, SFC, new Windows install, etc... will never fix this).

I would blame the RAM, except that you say you already swapped out the RAM. So,

1. Check cooling (fan speed, thermal paste, etc..)
2. See if there are firmware updates available
3. RAM errors were just reported in DIMM A2. So remove that DIMM and try running just on the B2 stick for a while
4. After that you might need to bite the bullet and swap the CPU & motherboard.

[Lang]

Hey David, thanks for the reply. I myself have also been pretty concerned with the temp of this CPU at times, it only ever rushes up to 95c then drops right back down. The "people" of Reddit claim that this specific chip is actually made specifically to go to 95c for max performance, anyways, I can't find anyone from AMD who is willing to back that up...

I will give the one stick a try, I have tried flopping them before, but that has never resolved it, so I'll be interested to see how one stick goes. As for firmware, MSI at one point sent me some super alpha channel BIOS to try, and even that didn't make a difference.

As for replacing the CPU & motherboard, I'm also at the point of being willing to do that (after fighting a $2700 custom built piece of metal that was supposed to be my life long dream machine for the last 7 months).




I have attached the log file in the attachments of this message.


Here is the contents of the .txt file
MemTest86 V10.7 Pro Build: 1000
PassMark Software
www.passmark.com

System Information
=============================================
System:
Manufacturer: Micro-Star International Co., Ltd.
Product Name: MS-7D75
Version: 1.0
Serial Number: To be filled by O.E.M.
BIOS:
Vendor: American Megatrends International, LLC.
Version: 1.E0
Release Date: 02/20/2024
Baseboard:
Manufacturer: Micro-Star International Co., Ltd.
Product Name: MAG B650 TOMAHAWK WIFI (MS-7D75)
Version: 1.0
Serial Number: 07D7511_N61E620538
CPU Type: AMD Ryzen 7 7700X 8-Core
CPU Clock: 4501 MHz [Turbo: 5476.1 MHz]
L1 Cache: 16 x 64K (307985 MB/s)
L2 Cache: 16 x 1024K (158576 MB/s)
L3 Cache: 1 x 32768K (121884 MB/s)
Memory:
Total Physical Memory: 65308M (52613 MB/s)
RAM Configuration: DDR5 4800MT/s / x2 Channel / 40-40-40-77 / 1.100V
Number of RAM SPDs detected: 2
SPD #0: 32GB DDR5 XMP 2Rx8 PC5-51200
Vendor Part Info: Corsair CMK64GX5M2B6400C32
JEDEC Profile: 4800MT/s 40-40-40-77 1.1V
XMP Profile 1: 6400MT/s 32-40-40-84 1.4V
SPD #1: 32GB DDR5 XMP 2Rx8 PC5-51200
Vendor Part Info: Corsair CMK64GX5M2B6400C32
JEDEC Profile: 4800MT/s 40-40-40-77 1.1V
XMP Profile 1: 6400MT/s 32-40-40-84 1.4V
Number of RAM slots: 4
Number of RAM modules: 2
DIMM A1: Empty
DIMM A2: 32GB DDR5 2Rx8 PC5-38400
Vendor Part Info: Unknown CMK64GX5M2B6400C32 00000000
SMBIOS Profile: 4800MT/s 1.1V
DIMM B1: Empty
DIMM B2: 32GB DDR5 2Rx8 PC5-38400
Vendor Part Info: Unknown CMK64GX5M2B6400C32 00000000
SMBIOS Profile: 4800MT/s 1.1V

[David (PassMark)]

I still think 95C is too high.
There is a lot of discussion about PBO / PPT setting for this CPU. Basically slightly underclocking it to keep the temperatures under control. So maybe that is something to look at.
Could do a temporary underclock to see if that fixes errors.

Would be super interesting to know the temperature of the liquid in the water cooler. Wonder if that is also near 95C. Do you get burnt if you touch the radiator?