My main PC is back!! Wheee!!!
Remember, it died on 2025-12-29. I'd say it was part of the "ASRock motherboards killing Ryzen 9800X3D CPUs" conspiracy, although I'm not discounting the possibility that some of the earlier CPUs had manufacturing issues or something. There are probably tens of thousands, if not hundreds of thousands of these CPUs out there, and only hundreds are reported to have failed, so I'd say I just drew one of the short straws.
One interesting thing about my case is that I'm a reliability nut. I never enabled PBO (Precision Boost Overclock). I never even enabled the normal boost function, so the peak frequency it ever saw was 4.7GHz (base frequency of 9800X3D). I even underclocked the RAM to 5200MT/s to lower the VSOC voltage. And it lasted 11 months or so.
Regardless of who's at fault (AMD or ASRock), it left a bad taste in my mouth, so I bought a new motherboard. Not too many choices out there supporting ECC memory, so ended up with an ASUS TUF GAMING B850M PLUS WIFI. I ended up with another NVMe slot, so I filled them all. This sytem has total of five SSDs (3 NVMe and 2 SATA). In the end, this board is way higher end than the one that killed the CPU (ASRock B850M-X WiFi), and I got a good deal on it ($170), so I'm happy with it.
Notice I don't like using motherboard supplied heatsinks. I prefer sticking heatsinks onto just the SSD controller chips (and not the NAND chips), and also leaving them in open air.
Anyway, the replacement CPU (brand new in box) came today, so I revived my main system. Dead CPU was 2452PGS (manufactured on the last week of 2024), and the new CPU is 2538PGE (manufactured mid-late September, 2025). Hopefully, if there were any early bird issues, this new one doesn't have it.
I've not had ANY PC component fail on me in MANY decades, until this one. I'm talking 45 AMD CPUs since I started keeping count sound time around year 2000. Oh well. Shit happens. At least my DRAM didn't get damaged.
In a few days, I'm going to write a novel length post on ASRock forum. Hehe.
