Restart randomly on AM5 chipset
Restart randomly on AM5 chipset
Recently installed a new AM5 setup for my partner and experienced unexpected restarts. The system can be down anywhere from five minutes to several hours of operation. It doesn’t crash or restart consistently; she can run multiple tasks and Blender renders smoothly for long periods, but eventually it fully crashes and restarts. I’ve only seen one BSOD, and these issues have persisted over weeks. Microcenter diagnosed it as a "failing CPU test." Despite this, temperatures, freezes, or lag haven’t been an issue. I’m struggling to find details about the problem and need guidance. Build specs: Ryzen 9 7900X, ASUS Prime 9070xt, OC Corsair Vengeance 32GB DDR5 6400 MSI Tomahawk MAG MAX, Wi-Fi 7, 7x670e Corsair, RM1000e, Windows 11 Montech, 360mm AIO. I’ve replaced components, done a MOBO CPU clean install, run MEMTEST, tested three separate GPU cables (two 8-pin PCIE, one 12VHP 8-pin PCIE), cleared CMOS, updated drivers and BIOS error codes, and used volmgr 162 kernel-power 41. I can attach dump files but haven’t been able to analyze them well.
I plan to replace the new PSU tonight to verify, though I suspect it would fail under stress or after prolonged use if it were defective. I downloaded one of the minidump files and sent it to ChatGPT for analysis, which reported a memory address overload issue—but I’m being cautious about that conclusion.
Do you really have DDR5-6400 memory with EXPO enabled? Most Ryzen 7000/9000 chips struggle to maintain full stability at that speed—DDR5-6000 remains the optimal choice. I suggest lowering the speed and checking performance. How many MemTest86 runs did you perform? Testing the entire 32GB would require several hours. You don’t need three separate PCI-E power cables; just use two. Connect the 8-pin and 8-pin in a daisy chain, using a single 8-pin connector.
I’ve set it to 6000 but it kept crashing at 5600. I ran MEMTEST and Windows Memory Diagnostic for several hours without any issues. It was using a daisy chain and had a separate setup, yet it still failed, so I decided to be cautious.
Start by resetting all changes in the BIOS if you made any adjustments. Verify for updates after touching anything, even if it doesn’t fully resolve the problem—it’s always wise to look. Try connecting the CPU to another machine; this might point to a PSU failure or bent motherboard pins. Inspect the motherboard carefully, though it’s unlikely. Based on Gemini Volmgr 162 and Kernel-Power 41, power issues could be involved, possibly related to power delivery problems. An experience on Reddit mentioned a similar issue with an Intel CPU, where switching the motherboard resolved it.
You didn’t necessarily end at the Gemini section—maybe you skipped ahead. I’m trying my best to be clear, though I might not have captured your full message. Also, I understand English can be tricky for some, and I can’t guarantee 100% accuracy.
I ran Memtest86 once more and saw over 10,000 errors on test 8. The computer started crashing each time I logged into Windows just before. I’m planning to replace two new memory sticks and try again tonight. It feels strange since I didn’t notice the issues before, but I’ll update the plan.