MSI MPG Z690 EDGE WIFI DDR4 XMP ISSUES

[TheK9]

For the longest time I have been doing stability tests using OCCT, and recently I noticed that stress testing my CPU, i9-12900k with XMP enabled on my 64GB ddr4 4x16GB kit it was spitting out hardware core errors. I figured out that disabling XMP resolved the issues. And after more digging, I discovered people have had a similar issue with the same board and CPU core errors with XMP. Someone suggested downgrading the bios in another forum post, so I did so, and low and behold it fixed the errors in OCCT, and the random program crashing. The version I downgraded to was 7D31v11. Through hours of testing, it seems ONLY this bios version allows me to run XMP at the rated speeds without program/game crashing, and hardware issues in OCCT. Is this a known issue? Is this being looked into? Are there any solutions besides sticking to this old bios version? This happens regardless of a CPU overclock of any kind, and XMP being the only factor.
Infomation about system:
CPU: i9-12900k
Motherboard: MSI MPG Z690 EDGE WIFI DDR4
Bios version pre downgrade: 7D31v14
Bios versions tried: 7D31v14, 7D31v13 , 7D31v12 , 7D31v11
Ram model: G.Skill Trident Z NEO Series 64GB (4 x 16GB) 288-Pin SDRAM (PC4-28800) DDR4 3600 CL16-19-19-39 1.35V Quad Channel Desktop Memory Model F4-3600C16Q-64G

 

[TheK9]

Update: as of more testing, v11 is now also producing the OCCT core errors

 

[plutomate]

Did you monitor your CPU temperature during the test? Since enabling XMP will most likely put more loading to the CPU/IMC. Right now it's hard to conclude whether it is a "CPU" or "Memory" related issue until you test it with memtest.
Any chance you can give link of other people having the same problem?

 

[TheK9]

Did you monitor your CPU temperature during the test? Since enabling XMP will most likely put more loading to the CPU/IMC. Right now it's hard to conclude whether it is a "CPU" or "Memory" related issue until you test it with memtest.
Any chance you can give link of other people having the same problem?

Temps are low, idling 30s, under load in 60s, under extreme AVX loads 80s. Weirdly enough, did a full memtest and got no errors, but OCCT still shows the core errors, and games and programs randomly crash to desktop. I've found a temp solution, by lowering my ram speed by -100mhz, from 3600 to 3500, but unsure if it'll permanently fix this. Oh, and a side note, the ram works fine in another system, i9-9900k.If I come across the posts again with people experiencing similar issues ill post them, but its been a few days and I'm struggling to trace them back as my history auto clears after a day.

 

[Ezmafl]

For the sake of it, have you tried removing two sticks and seeing if the errors still occur?

 

[raimondkieftenbel157602de]

For the sake of it, have you tried removing two sticks and seeing if the errors still occur?

This, try this, 4x16gb is pretty thought for the Memorycontroller off the cpu and could need some tweaking

Also it could be the case,you cant get it stable at XMP settings(because off memory controller limitations).I would aim first at an lower speed and check if that is stable.

I really dont think its a bios thing. I got the same Board,but its stable with with XMP with all bios versions,with tre kit I use.

 

[dagedy]

IMC in your CPU have much to do with your spec. To get stability you should try manually increase SA Voltage and, DRAM Voltage.
Try start from SA Voltage, set 1.25 - 1.3V then run OCCT