Multiple WHEA Errors at stock (x870e Carbon + 9800x3d)

clement.fg153702d7

New member
Joined
Feb 25, 2025
Messages
3
Looking for insight as to which part might be faulty or if it might just be a driver issue/Bios firmware problem?

Description of Original Problem:
Receiving a lot of WHEA errors when running OCCT on my new PC build with no overclocking. Running a 3 minutes CPU+Mem test or Mem test will give 90+ WHEA errors. This is without even the EXPO profile. I have a RAM kit rated for 6400mhz but even at 5600mhz is already unstable.

Troubleshooting:
OCCT runs CPU test without errors
OCCT gets errors in CPU+Memory test
OCCT gets errors in Memory test
Ran MEMTEST86, completed with zero errors
Flashed to 3 different BIOS revisions, including latest and oldest.
Updated to latest chipset driver
Confirmed no overclock (not even RAM Expo profile or PBO)
Cleared CMOS
Tried running test on each stick individually with only 1 stick inserted, still same issue
Tried different RAM sticks: same issue
Checked temps: temps are great
Tried resetting BIOS to "Optimised Defaults" to make sure no OC
Updated Windows and GPU driver
Tried clean windows install on different nvme drive and slot

Computer Type: Desktop
GPU: ASUS TUF RTX 3090
CPU: RYZEN 7 9800X3D
Motherboard: MSI MPG X870E Carbon Wifi
BIOS Version: 1.a23 (most recent)
AMD AGESA Version: ComboAm5PI 1.2.0.3a Patch A
RAM: 2x24Gb Trident Z5 Neo (Currently not using Expo profile, so 5600mhz)
PSU: MSI MEG Ai1300P PCIE5
Case: Phanteks NV7 Black
Operating System & Version: WINDOWS 11 Pro 24H2 , OS Build: 26100.3194
GPU Drivers: GEFORCE GAME READY DRIVER - 566.36 Chipset Drivers: AMD X870E (Promontory PROM21L.7)
 
Can you run TestMem5 with the 1usmusv3 profile (see point 5 of my RAM thread)? I trust that the most, regarding RAM stability testing. If OCCT shows a bunch of errors right away, then TM5 should also show an error in record time, i suppose.

Then i would normally assume that the BIOS / memory system has an aversion against this RAM, and different RAM might work better, but you said you already tested a different kit? Which one exactly?

I guess it's also possible that the CPU's memory controller has some problem/defect. Wouldn't be the first AM5 CPU that was partly defective. But that's difficult to test without using another AM5 CPU. One thing you could perhaps do, check for bent pins in the CPU socket. But yeah, to me, with this kind of issue, the CPU (in particular, its IMC) would actually be more likely to cause it than the board. But it would need further testing.
 
Can you run TestMem5 with the 1usmusv3 profile (see point 5 of my RAM thread)? I trust that the most, regarding RAM stability testing. If OCCT shows a bunch of errors right away, then TM5 should also show an error in record time, i suppose.

Then i would normally assume that the BIOS / memory system has an aversion against this RAM, and different RAM might work better, but you said you already tested a different kit? Which one exactly?

I guess it's also possible that the CPU's memory controller has some problem/defect. Wouldn't be the first AM5 CPU that was partly defective. But that's difficult to test without using another AM5 CPU. One thing you could perhaps do, check for bent pins in the CPU socket. But yeah, to me, with this kind of issue, the CPU (in particular, its IMC) would actually be more likely to cause it than the board. But it would need further testing.
I ran full 10-pass MemTest86 and didn't get any errors, and I did test with another RAM kit, a very basic corsair ddr5 ram kit (4800mhz 2x16gb) that I was using on my previous build. I'm sending the board tomorrow for RMA and will exchange the CPU for a new one since I got that one in-store and it's still within the 30 days return warranty :/
 
TM5 is more detailed than Memtest86. MT86 is good for testing if a module is defective, and for basic stability, but it's not the last word on stability. That's why i suggested TM5.

Don't RMA the board. If the socket pins are all 100% ok, then the board has a lower chance of causing this, compared to the CPU. All the "brains" of the memory system is in the CPU, in form of the IMC (integrated memory controller). The board has more of a passive role in the memory system. That's not to say nothing can go wrong with a board regarding the memory system, however, to RMA the board at this point might prove to be a waste of time and resources.
 
Looking for insight as to which part might be faulty or if it might just be a driver issue/Bios firmware problem?

Description of Original Problem:
Receiving a lot of WHEA errors when running OCCT on my new PC build with no overclocking. Running a 3 minutes CPU+Mem test or Mem test will give 90+ WHEA errors. This is without even the EXPO profile. I have a RAM kit rated for 6400mhz but even at 5600mhz is already unstable.

Troubleshooting:
OCCT runs CPU test without errors
OCCT gets errors in CPU+Memory test
OCCT gets errors in Memory test
Ran MEMTEST86, completed with zero errors
Flashed to 3 different BIOS revisions, including latest and oldest.
Updated to latest chipset driver
Confirmed no overclock (not even RAM Expo profile or PBO)
Cleared CMOS
Tried running test on each stick individually with only 1 stick inserted, still same issue
Tried different RAM sticks: same issue
Checked temps: temps are great
Tried resetting BIOS to "Optimised Defaults" to make sure no OC
Updated Windows and GPU driver
Tried clean windows install on different nvme drive and slot

Computer Type: Desktop
GPU: ASUS TUF RTX 3090
CPU: RYZEN 7 9800X3D
Motherboard: MSI MPG X870E Carbon Wifi
BIOS Version: 1.a23 (most recent)
AMD AGESA Version: ComboAm5PI 1.2.0.3a Patch A
RAM: 2x24Gb Trident Z5 Neo (Currently not using Expo profile, so 5600mhz)
PSU: MSI MEG Ai1300P PCIE5
Case: Phanteks NV7 Black
Operating System & Version: WINDOWS 11 Pro 24H2 , OS Build: 26100.3194
GPU Drivers: GEFORCE GAME READY DRIVER - 566.36 Chipset Drivers: AMD X870E (Promontory PROM21L.7)
hi im getting the same exact issue as you, did you fix it?
 
Back
Top