Trident-z on x570 configuration experiencing several Memtest86 failures and BSOD occurrences.
Trident-z on x570 configuration experiencing several Memtest86 failures and BSOD occurrences.
Hello, I understand you're seeking assistance with diagnosing issues on your RAM following several errors during Memtest86. The system includes: Trident Z Neo DDR4-3200 CL16-18-18-38 1.35V, 64GB (2x32GB) F4-3200C16D-64GTZN, built with Gigabyte X570 AORUS PRO WIFI (rev. 1.0), AMD Ryzen 9 3900X, an XFX XTR2 850W power supply at 80+Gold A, and an RTX 2060S along with a couple of NVMe drives. More details can be found here: [link].
You've collected dump files from the latest GSODs, a Memtest86 log, error screenshots, and some videos. After enabling "Memory Integrity" and Hyper-V, you experienced multiple BSODs and bugs, possibly linked to recent Nvidia driver updates and Windows patches. There were also issues with micro mouse stutters and persistent errors on both Windows Memory Diagnose and HammerTest.
One stick displayed its name and specifications correctly, while the other showed capacity and speed but failed Memtest86 recognition. It took over 12 hours and reported 7 errors. You're wondering if swapping RAM sticks could help, fearing damage to either the RAM or slots, especially since you're concerned about warranty implications.
Other tests like HCI and additional Memtest86+ might be useful. I’d appreciate any guidance on safely checking the RAM without risking further damage, as well as confirming whether a faulty stick is the issue. Your caution is wise—especially with the warranty and potential risks involved.
This situation might not improve things. The rev. 1.0 boards on X570 had a poor memory setup that performed poorly, and because they were only available for a short time early on the X570's life, they received minimal long-term BIOS updates. Memory support tended to decline with newer BIOS versions, making it harder to test stability. For 3200MT/s speeds, it shouldn't be a major problem, but issues become noticeable above 3733MT/s. Rolling back to an older BIOS version could help Memtest detect problems, though the chances are low. It's unlikely the memory itself is faulty, but testing all slots might reveal improvements. Memtest86 is effective at spotting errors, so any instability would point to the CPU controller or RAM rather than the sticks themselves.
This forum is really great! I updated my bios a few weeks back and discovered the newer versions are mainly about memory architecture updates. Your point makes perfect sense now. I plan to revert the bios, probably to the version with at least Age 1.2 so TPM works with Windows 11. I haven’t rolled back yet but might. For now, after some discussions I set my voltage to 1.35L and XMP High-frequency at level 1. I’ve been testing Test8, which could finish in about 15 minutes and show 7 errors. Once I adjust it, a Passed test for the first time would be amazing. It’s not much yet but it gives me hope and relief. The full test took over 13 hours before... If things work out, please let me know if there’s anything else I should do. I’m unsure if manually tweaking indicates CPU or motherboard issues, but I’d appreciate any tips on memory timings. Thanks!
You'd likely prefer a release within the first 6 to 12 months after the board launch, as support from Gigabyte would have diminished by then. Windows 11 doesn't require TPM for basic operation, which makes it unnecessary on a typical home setup—especially if you already have it enabled by default. Keep in mind that AMD's changes to PBO in later AGESA versions could improve performance slightly on older BIOS versions, though only benchmarks would confirm this. You should only consider upgrading your BIOS if stress tests indicate failure or if you're experiencing crashes, not just for general performance gains. The only practical reference I have is Buildzoid's memory overclocking videos, which explain timing details in depth but are quite intense and not suited for all viewers.