Irregular crashes, highly volatile machine (13900K + RTX 4090)
Irregular crashes, highly volatile machine (13900K + RTX 4090)
I’ll begin by testing with one RAM stick at a time during regular use and observe the results. With three crashes today, I expect to quickly identify any issues. Regarding your question—should I replace the RAM now or wait for the retailer’s feedback on the CPU? I’d rather avoid random swaps unless there’s a solid reason. I’ll also make sure to send along any new dump files that appear during testing, as they could provide more clarity.
Not every 13th and 14th generation CPU is impacted by this issue. The i3 and non-K i5 models remain stable according to Intel. Many of these chips are actually from the 12th generation, just rebranded recently. I own a 13400 processor, and when I purchased it in late 2023, the problem hadn't been disclosed yet, so I might have chosen a more robust alternative then. Now it seems like I avoided a serious problem by chance.
Two examples display an NMI being transmitted. NMI stands for Non-Maskable Interrupt, a special type of interrupt that forces the CPU to pause and address it right away, bypassing the normal execution queue. This typically indicates a serious hardware fault. Memory generally can't generate NMIs unless ECC is enabled, making this situation more likely related to the CPU itself on consumer systems. This strengthens the suspicion of a hardware problem rather than a software issue. We don’t know the exact reason for the NMI trigger or what caused it, but we see an NMI being directed to the BOSD.
That’s really interesting – thanks for pointing that out. I wasn’t aware of that detail about NMIs. Definitely adds more weight to the CPU theory. I'm currently waiting to hear back from the retailer about a potential RMA, so fingers crossed. Furthermore, I heard back from ASUS support. According to them, the BSODs I’m seeing (KMODE_EXCEPTION_NOT_HANDLED, PAGE_FAULT_IN_NONPAGED_AREA, KERNEL_SECURITY_CHECK_FAILURE) could be caused by my memory kit – G.Skill F5-6000J3038F16GX2-TZ5N – which, apparently, is not listed as compatible with my motherboard and CPU combo on their QVL. They basically said that even if the RAM seems to work fine at first, issues can crop up later in unpredictable ways and become increasingly hard to diagnose. And that they won’t really proceed with further support until I’ve tested with a QVL-listed kit. So... I’ve now ordered a compatible RAM kit from their list to rule that out once and for all. I’m planning to sell the old kit on eBay. To be honest, I still think the root issue might lie elsewhere, but I get why they want to eliminate memory compatibility first. Makes sense from a support perspective. Once the new RAM arrives and I’ve tested with it, I’ll share another update — and more dumps if it still crashes.
Consider the possibility that the issue might be with the RAM rather than the CPU. It could send corrupted signals to other parts, which often appear as error messages. If it seems faulty, an RMA might be the best option.