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?
Issues likely stem from memory instability, possibly due to corruption. Despite stable tests in a fresh Windows install, problems like unhandled exceptions persist. Recent RAM upgrades haven’t fixed the matter. System runs smoothly for now but crashes increasingly often, especially with games and Discord. Memory issues seem to be the core problem. I’m unsure what to replace next without more testing.
Intel provides a tool to verify your CPU. Test with one RAM stick (no XMP enabled) and observe the results. Provide feedback.
Received. Is this assessment sufficient to identify issues with the CPU's memory controller? --- IPDT64 - Revision: 4.1.7.39 --- Started on 4/26/2023 at 3:05:20 PM --------------------------------------------- Testing CPU 1 results: All tests passed. CPU 1 passed every check including Cache, MMXSSE, IMC, Prime Number, Floating Point, Math, GPUStressW, CPULoad, and CPUFreq. IPDT64 Passed --- End of test at 3:09:08 PM
Not familiar with any videos explaining this, but the situation arises when RAM instability occurs. Windows files might get corrupted because the operating system moves data around during background operations that programs depend on. This problem usually disappears after a clean installation if the RAM is functioning correctly. People often mention Prime95 Large FFTs, though it seems each stick passing standard tests suggests the hardware itself may be faulty rather than the memory. This behavior isn’t uncommon with DDR5, as the power management circuitry can overheat and reach temperatures above 50-60°C depending on case ventilation. Most DDR5 components are temperature-sensitive, but they’re rated for up to around 80°C under JEDEC conditions, so running below that should be fine. Given your thorough testing of memory, a memory fault seems unlikely. The main suspects are the CPU, motherboard, or SSD. Trying a spare SSD to boot Windows could help—if it works, you’re off the hook. If not, consider other troubleshooting steps. For motherboard issues, outdated BIOS revisions can cause instability; switching to another board might resolve it. If the problem persists, overvolting the CPU slightly and running stress tests could help. Adding a small positive voltage offset (50mV) is another option to see if it stabilizes performance.
I didn't consider the hard drive. Actually, I have two more NVME drives, so I can install a new Windows setup on one of them and check it out. (Having an 850SN 4 TB as the OS drive was great, you know.) I plan to install it tonight and will follow up. I'm mainly concerned about applying voltage—not because of overheating, but because the chip should stay stable without any changes. It's watercooled and temperatures are normal. For RMA, I need to explain this board since I'm leaning toward it. It's been a while since I filed an RMA before; usually it was just "It's dead." I just don't want to be caught wondering if the issue is with the component or something else like the CPU or memory. I'm going to ask a local shop if they'll let me pay 50-100 bucks to borrow a test bench to verify the RAM and CPU are working properly.
Absolutely, the extra 50mV is probably safe and just helps with troubleshooting. If raising the voltage stops the crashes, return it to Intel with notes about stress tests. Avoid using it for more than a day unless necessary. For unstable systems, a day should suffice to verify functionality. This process can be challenging, but luckily I've never had to replace a motherboard. Often, when boards cause random blue screen issues, they get returned and tested. If no problems are found, they send a replacement. The exact outcome may differ. If possible, consider this a solid solution. Be sure to mention it so we can notify you.
I only had around 20 minutes before falling asleep. I installed Windows on a fresh USB drive and moved it to another M2 drive in my rig. So far, the errors aren’t showing up, but I’m still cautious. I plan to keep it running all night downloading something. Besides checking Discord, Chrome, and the Epic Games Store, are there other stress tests I could run? A dead M2 drive seems like a straightforward fix if that’s the issue. Crystal Disk is usually the best tool for confirming the drive problem too.
It's a solid choice, though it has its flaws. I've noticed many failing drives often don't report errors there. Feel free to run some checks, but if the problem disappeared after replacing the SSD, it's likely the drive itself was faulty. Start by inspecting the SSD first—it's usually the simplest fix. Consider additional stress tests like Linpack or Prime95, but the most reliable method is to let the system operate normally and monitor for any issues.
I'm going to shut down now, but I left the other installation on the suspect drive. Tomorrow morning I'll attempt to boot into it, run various programs, and then switch to the safe drive to check for any clear differences.
It doesn’t look like the SSD is the issue—I’m still encountering the same errors. Unless both the Epic Games Store launcher and Discord have a temporary glitch affecting my setup, there’s likely another problem. I’m pretty confident it’s related to the motherboard. I ran some benchmarks in January to establish a baseline, but it wasn’t continuous testing (I didn’t run it nonstop). Any other suggestions?