Occasional freezes, needing a restart
Occasional freezes, needing a restart
Over the past few weeks I experience unexpected freezes that require a full system restart. The displays become unresponsive without any visual error, only background noise remains audible. If the music plays, I can still hear it. The mouse and keyboard stop working completely. Occasionally it seems to recover within a minute, but usually I have to start over. At first I dismissed these issues because they were infrequent (about once a week or so, sometimes with longer gaps). Recently they’ve increased in frequency—possibly daily—and I began investigating. It likely started when ads on Twitch or YouTube loaded, especially during game transitions. Now, since I no longer watch ads on those sites, the problem appears to occur while gaming. The first incident happened during a Twitch/YouTube ad load while I was playing; other times I had multiple applications open (streaming and several Chrome tabs). Recently, after switching my PSU and using a better one, the crashes intensified. I reset four times in a single League of Legends match, so it clearly worsened. It doesn’t happen during stress tests or when I play Sekiro. I haven’t tried other games yet, but I’m ready to test them if needed.
My hardware: CPU Ryzen 3600 (original), GPU AMD 6600xt Sapphire Nitro+ (used for casual mining), 32 GB RAM, Goodram IRDM Pro Deep Black 3600mhz cl18 (XMP enabled), Windows 11 64 bit (from Windows 10 yesterday). System runs on NVMe WD SN570 (plus several older SSDs and HDDs). BIOS: MOBO Asus ROG STRIX B550-A. Power supply: Corsair RM650x (previously a 7-year-old SilentiumPC Vero M1 600W, now well-cooled).
I’ve tried several fixes: reinstalling Windows, updating drivers and games (though some older SSDs and HDDs remain), running Memtest with XMP disabled for weeks, using Prime95 blend tests, Linpack, Memory, VRAM, Power, CPU stress tests, and even changing thermal paste. I also replaced my PSU to a higher-tier model.
I previously adjusted XMP settings using Ryzen memory tools, which helped during the first few crashes. After a period of inactivity, I resumed XMP and it worked briefly. However, I’m unsure if that was a temporary fix or just a false sense of security.
Recently, after switching my PSU, I ran Prime95 for an entire night, then turned off XMP to see the outcome. The results were unclear, but I feel completely uncertain now. I suspect the power supply might have been the real culprit, not the PSU itself.
Consider turning off MultiPlane Overlay MPO. Also, update the latest BIOS version. Regarding chipset drivers, did you install the newest ones from AMD that resolve black screens and some BSODs? Are you using them? Also, what does Thaiphoon Burner say about your RAM? If you installed a Dram tuner for Ryzen by 1usmus, you should have checked the exact RAM model.
Thanks for your reply! Since the previous update: I turned off MPO and updated the BIOS to the newest version. The CPU ran smoothly through Prime95 Blend for eight hours without any issues or freezes. I was already using the latest GPU drivers (version 22.11.2) and recently downloaded the newest chipset drivers from AMD, specifically the ones designed for my motherboard. I enabled CPU virtualization to support Thaiphoon on Windows 11, just letting you know. After that, it crashed again after a few hours. Now I plan to test auto RAM settings. Both kits are identical, and Thaiphoon doesn’t recognize the serial number, only the manufacturing date is the same: so it’s Hynix CJR / c-die.
Is there a VBios update available for the GPU? Have you verified the condition of your drives using tools like Hard Disk Sentinel? Are your SSDs running the most recent firmware versions? Prime95 Blend also puts pressure on the RAM, so based on what you did, it seems unlikely the issue is with the RAM. Do your 3DMark Firestrike and TimeSpy results match those achieved by users with comparable hardware? MSI Kombustor or Furmark are functioning properly, with no crashes appearing after just 1-2 hours of use?