The concerning question is: Is the issue linked to memory, the motherboard, or the CPU memory controller?
The concerning question is: Is the issue linked to memory, the motherboard, or the CPU memory controller?
It seems there might be some confusion. The issue isn't necessarily about the motherboard being faulty, but rather about unusual problems with your system. You could reach out to ASRock to report these concerns and request a replacement.
I’m planning to proceed with that. However, I wanted to ask if it’s worth considering an ECC board, such as the w680. From what I’ve learned, this might be the only platform supporting ECC on DDR5. Maybe this experience has made me cautious, so I’m trying to figure out how to avoid that. I’m not sure exactly how it would fail, but I’m worried about instability. A w680 motherboard could potentially fail in some way, though I’m not confident it would affect the hardware itself. I’d prefer to understand what’s causing these issues before making a change. Most importantly, I want to be sure that running 128GB DDR5 isn’t causing problems with memory controls. If that’s true, I’d feel reassured and would likely replace the board when I get a new one from Asrock.
No, it doesn't harm the board. If it did, I'd face many more dead motherboards (I really enjoy memory overclocking, where you often need to tweak timings until they're stable). It could ruin the operating system and possibly corrupt the BIOS in extreme cases (I've seen this happen), though these issues can usually be resolved with a Windows reinstall and a BIOS update. AFAIK AM5 supports ECC DDR5, although finding kits is tough. Adding ECC isn't necessary unless you're using 128GB of RAM, but the price for unbuffered ECC RAM is usually very steep—potentially doubling your current RAM cost. If you can afford it and manage to install proper VRM cooling for a W680 board (most I know have terrible VRMs that throttle high clocks), it's a reasonable approach.
You might need to consider a VRM waterblock to ensure proper cooling for your motherboard. The custom loop cooling alone may not provide sufficient heat dissipation.
I forgot this board existed, this should be OK with a 13900k as long as there's a bit of airflow over the VRM. https://www.asus.com/motherboards-compon...-ace-ipmi/ Most of the W680 boards look closer to this monstrosity and do actually need VRM water cooling to have a shot of powering a 13900K without power throttling: https://www.newegg.com/supermicro-mbd-x1...ption=w680 motherboard&cm_re=w680_motherboard-_-13-183-808-_-Product Also, a custom loop counter-intuitively actually has a VRM run hotter because it's less likely to get direct airflow like it would from a CPU's air cooler. It's not like that matters for the vast majority of high end boards these days, the Taichi you're currently running could be run without a heatsink with zero airflow and should still be OK because the VRM is so overkill, but it will technically run hotter with a custom loop than an air cooler.
Just an update - I've received the motherboard and new ECC ram. I installed them and reinstalled windows. I am STILL getting the same errors. This is fucking nuts. Ran the CPU through Prime Blend for 30 minutes no errors - also ran it through the OCCT tool for the large/small AVX2 and a large SSE or whatever they are called - no obvious errors - reported a minor PCI error from event viewer that shows a ID of 17 with a corrupted. I have no idea what to do now - if there is a problem with the memory controller in this CPU, why can't I get it to manifest during a stress test - I can't even open up my unreal engine project file
Before swapping the motherboard, I reset the CMOS and reinitialized the BIOS. The only adjustments were restoring power after an AC outage. I used the Intel diagnostic tool on that machine. After replacing the board, the same issues persisted. The BIOS was original, and even after a single day, I couldn’t recall any significant changes. I haven’t run the Intel tester yet. The event viewer doesn’t display anything relevant to these crashes, especially during the incidents. I had a secondary video card for a single display that I plan to remove once home to rule out a memory access problem. This error is really puzzling. I downloaded a fresh copy of The Unreal Editor and it’s also failing. I need to verify if corrupted project files are the cause—this file appeared to be a core engine asset, likely new since the download from the Unreal server directly on this PC.
Additionally, it's important to observe that every retrieved memory address returns null. During crashes, it seems linked to missing pointers.
Transfer the installation to a USB drive using a LiveUSB copy of Linux Mint (similar to Windows). Keep it running from the USB for some time. Linux is very sensitive to poor hardware. Any issues suggest a software problem. If it fails, it points to faulty hardware.