After hearing how many 9XXX CPU’s have been blown up by the newest ASRock motherboards, I’m worried that it’s only a matter of time till it happens to me.
My RIG
CPU : 9800X3D
MOBO: ASRock x870e Taichi Lite
RAM: G.SKILL Trident Z5 Neo RGB 64GB (2x32GB) DDR5 6000MHz CL30
STORAGE: 1TB SSD WD Blue
PSU: 1200W 80+ Gold Be Quiet Dark Pro
Background
This isn’t my first PC build. As of today everything works, but I haven’t updated the BIOS to the latest version 3.20. I’m on 3.14
The computer was on for maybe 50 hours tops. There have been some strange behaviors such as the board giving me a 31 error code indicating bad RAM. That was fixed by flipping the power button on the PSU, waiting 1 min and powering the PC back on.
I’m actually contemplating getting an affordable B650 mobo from Asus or any other brand besides ASRock. Is that necessary at this point, or am I just overreacting?
All comments are welcome.
Thank you for your time.
Huh, 157 is quite a bit… It’s still a small percentage but it’s atypical for CPUs to outright fail unless overclocked aggressively.
I probably wouldn’t. It’s hard to calculate odds from 157 anecdotal cases (not everyone posts about something like this – and who knows how many 9800x3d are out there), but it is for sure less than a percent. And if it does fail warranty should back you up. So unless it’s some scenario where your income depends on the system it’s not rational to spend a couple hundred to mitigate a (much smaller than) 1% chance of something happening.
Yup. And some of what’s being referred to as death is an ASRock BIOS bug which blocks post. That’s what the 3.20 BIOSes are supposed to fix and there’s plenty of reports 3.20 flashback does, indeed, cause some of the “dead” systems to post normally. But there’s reports 3.20 didn’t have any effect and lots of swapping in a different processor working fine. So it looks to me like Granite Ridge X3D probably has an early failure mode that AMD needs to sort out. (Or maybe it’s fixed but hasn’t saturated in channel yet.)
There’s at least a single digit number of cases of things getting in the socket or bent pins too. Those are presumably background failure rate getting rolled into the two or more other issues here.
Yeah, same. Asus, Gigabyte, and MSI all have failures too so it’s not like changing the board is a fix. It does seem to improve the odds by more than can be explained from just 3.20, so there might also be something with ASRock’s loadlines that makes X3D more likely to fail (or made, as @maker’s suggested).
If the 9900X or 9950X I’ve built quit I’ll RMA but for now there’s no meaningful action I can see to take. One reason I bumped them to 3.20 promptly is just in case it’s some kind of degradation thing.