Need assistance figuring out a strange BSOD problem
Need assistance figuring out a strange BSOD problem
This looks like memory from the dump files. Memory doesn't have to mean RAM, but it's usually the main suspect. Windows puts low priority data from RAM into the page file and loads it back in when needed so storage can look like memory (And memory can look like storage). The memory controller is in the CPU and if this fails it will just look like memory. When it's storage about half of the dumps will usually blame storage or storage drivers, which I don't see here, so it's likely not storage. If anything is overclocked or undervolted, remove it. Make sure that Precision Boost Overdrive (PBO) is set as Disabled in the BIOS. To test the RAM, use the machine normally with one stick at a time. If just one of the sticks cause crashes, faulty stick. If it crashes with either stick it's probably the CPU. Memory testers miss faulty RAM fairly often with DDR4 and newer so I don't trust them. Lastly, not sure if you were experimenting with RAM speeds, but in at least one, the RAM speed was 2133MT/s (The CMK32GX4M2E3200C16 kit) . In the others it wasn't included (Not your fault, Windows just does that sometimes). If you haven't played around with it, but you know that you set it up, it could indicate a RAM issue. If the motherboard detects it as not stable, it will reset the speed.
The situation might be anything but I experienced something comparable twice on different platforms. First it was related to the RAM (Corsair Vengeance naturally) Then the PSU ("direct power") Also a Kingston SSD caused issues. Short version: you usually have to replace components until you identify the problem. Also check if Windows and BIOS are set to default. If not, consider a CMOS reset and a fresh OS install (back up your data first).
I understand your point, but just using content from Amazon isn't really a good idea... A fresh Windows 10 installation would be better since your system is built for it. Also, did you ever upgrade to Windows 11? Or were you simply updating from Windows 10? Honestly, getting new RAM and a Trident Z would be a better choice—those Corsair kits didn't work well with Ryzen 3000 CPUs. Since I recently got G-Skill, everything runs smoothly without crashes or BSODs. (Just a heads-up, the sticker matters especially for older Ryzen models.)
I recently purchased a 1000w Lian Li PSU and received it a few days ago. While drafting this note, it arrived promptly. I completed all cabling and still encountered a BSOD during first boot due to APC_INDEX_MISMATCH while launching And. For the Samsung 990 model, I immediately installed Magician upon buying it to perform an update—it's now fully updated. Over the past two years, I've conducted tests without any block failures appearing. The system reports 93% health with 105TB of write activity, and SMART status shows no red flags. Here are my findings: I ran the built-in MSI SSD testing tool from the BIOS and used Samsung Magician multiple times to troubleshoot. Today, after an extended diagnostic with Magician, everything came back clean. Could a faulty DIMM slot be causing this issue? I considered testing one stick in a different slot each time to eliminate that possibility. Since then, I replaced those two Vengeance RGB Pro RAM sticks with brand new Vengeance LPX units. Despite these changes, the same problems persist. I suspect RAM might not be the issue—unless I'm unlucky and experienced issues with RAM purchased years apart. If it's the CPU, I don't have another AM4 board to test. It seems like I'm facing a situation similar to a "lemon CPU" or possibly my motherboard. I've replaced all components except the SSD, which I've already verified.
It's relatively rare, but the challenge lies in the fact that certain boards dislike using the non-recommended slot. If a slot other than A2 seems problematic, it's hard to tell if the issue is with the slot itself or with the board's preference for using that slot alone. I understand this point. If it's helpful, I've encountered many defective Zen 2 CPUs from the 3000 and 4000 series. Zen 2 tends to behave oddly, often appearing as memory regardless of configuration. I haven't seen many 3800x models, mainly the 3600/x and 4600H/4800H variants. The 3600 model performed exceptionally well, which might explain its popularity. Before proceeding, let me emphasize that this isn't about storage performance. My frustration stems from the way NVMe was handled—SMART was significantly reduced with NVMe support. The percentage values aren't linked to current health; they relate more to wear metrics, especially remaining warranty writes. Health assessments have been inconsistent for over 15 years across all drive types since manufacturers decide when a drive should fail before reporting status changes. In the upper half of CDI, the only meaningful data is the temperature reading. Regarding parameters, this is where NVMe drives were heavily compromised. Almost all useful info was stripped away, leaving just Critical Warning and Media/Data Integrity Errors. I'm confident I've encountered over a thousand faulty NVMe units, with fewer than five showing any recorded issues in these metrics—or any other indicators. I view SMART as largely useless for NVMe drives outside of temperature readings.
It seems the SSD might still be functional. You could try purchasing a replacement AM4 CPU, test it, and send it back for a refund through your nearby computer parts retailer or Amazon. Just confirm their return policy before ordering.
Thanks for the valuable feedback. I'm hoping the 3600 I just purchased will perform well right away—maybe it's a good sign! I've had some concerns about SMART readings on modern SSDs before, especially with how different they are from traditional hard drives. I'll remember that. The Ryzen 5 3600 I ordered from Amazon should arrive tomorrow or so. Appreciate all the guidance; I'm gaining a lot of knowledge. I'll keep you posted once I have the CPU and start testing. My weekend is busy, so it might take a few days before I can share my findings.
The situation is that SATA SSDs still rely on the "old" SMART features, similar to HDDs. For a while, new drives kept using the same outdated parameters as NVMe ones. I think I might have seen some early NVMe models with the older SMART, but it could just be my imagination. If the latest SMART is simply what they need for modern tech, why would SATA SSDs adopt the updated reporting? And since flash storage is what SMART should track, shouldn’t new systems also monitor things like reallocated sectors, pending sectors, and uncorrectable errors? It seems a bit confusing to me. These questions aren’t meant to be taken personally—they just feel like frustration directed at clouds.
I observed that... Believe it's just some window tricks, but I see...