That Windows 11 update kept killing a shop PC at 2am, so I checked the RAM sticks one at a time
A customer's Ryzen 5 machine kept blue screening only during overnight backups. I pulled one 16GB stick, ran memtest86 for 4 passes, and it came back clean, but the other stick threw errors in about 20 minutes. Turns out the bad stick was in slot A2 the whole time and only flaked out once the board warmed up. Anyone else see memory go bad only under load like that?
That A2 slot getting flaky once things warm up is a classic heat soak issue. Sticks can pass cold and then fail when the board hits 40C or so, which is why overnight backups catch it. Good call testing them one at a time instead of trusting a single clean pass.