原文
UPDATE:
AMD was quick to respond and provide a pre-paid shipping label for the RMA. I also heard from ASUS a little latter, who also offered replacements. It just came a little to late as to reduce downtime and for some stellar analysis content I had already sent them off to Steve. 🙏 Bless. No shade to ASUS though, It's only been 1 business day and this occurred end of week. Both AMD and ASUS have done a fantastic job in trying to make this right. I do feel for AMD and ASUS here as they did request that the parts be sent to them. It would have been good to have them get it first to provide an official solution quicker. Sorry guys.
Here are some images I have taken with magnifiers highlighting the damage: https://photos.app.goo.gl/tUKbc3tDWL6B6qqZ9
Comment image
Here is the full system: https://ca.pcpartpicker.com/list/bdXJhk
As for timeline and some additional information...
7600X has been running without issue since mid February.
7800X3D was installed and running 24/7 since the date of purchased (Sunday, 04/16/23)
No CPU overclocking, just memory via the EXPO 1 profile.
F5-6000J3038F16GX2-TZ5NR is on the QVL.
No issues with temps or performance to date.
Have been gaming a decent amount on the system. Got 30% gains in ARMA 3 (Very CPU heavy game), MSFS both VR (2x framerate from 30FPS -> 60FPS) and non-VR (30FPS with no FPS improvement over 7600X), and various other games both VR and non-VR. CoD, Hunt Showdown, Rocket League & Breachers to name a few.
Returned home after leaving my system in a typical idle state with nothing strenuous running (I don't dismiss the fact that a background task could have been running) to the system unable to POST and a QCode of 00. Of which is typically associated with no CPU.
The AIO was hella toasty upon removal, will burn you hot, specifically the bracket. I speculate the short to have caused the heat. Especially if it continued to short while on. No form of protection was tripped.
The pins on the LGA do look intact, just recessed and slightly charred.
I did attempt to update the BIOS to 1101 from 0922 during installation, however due to issues with EXPO profiles being unstable and not POSTing, I returned to 0922. 1101 states "Improve system performance and stabilize AMD Ryzen 7000 X3D series processors", not that the your CPU and MOBO die and become murder and mystery.
The last point is my main suspicion, and that the BIOS update needs to be made mandatory. The sad part is, if you do that EXPO profiles will not work for this QVL memory, but it is better than frying everything. This is speculation. I think it is reasonable to think that not updating ones BIOS should not do this. Sure recommended, but not necessary. If so, it might be time for the industry to do BIOS updates without a CPU, with versions only supporting a qualified list.
My memory, F5-6000J3038F16GX2-TZ5NR is on the QVL. But as previously mentioned, would not POST with the EXPO profile from the Ryzen X3D "stabilizing" BIOS version. After reading that others were just reverting to older versions when EXPO wasn't working claiming instability, I did the same. I believe these posts were prior/not on X3D processors.
EDIT: After speaking with ASUS, it was brought to my attention that this technically isn't warrantied with AMD due to the use of EXPO. Which was news to me. The board is though through ASUS. I have always thought of EXPO/XMP as "verified/approved" settings for memory, and not so much an overclock. Especially since the QVL reflects that. This is not the case, and in fact the rated settings also vary by CPU, so just because the QVL says 6000/30-38-38-96, does not mean that it should work across CPUs. With this in mind, as I state above, I purposely went back to a BIOS version which POSTed with EXPO which was not an approved version for X3D series CPUs and chalked this up to BIOS instability rather than a potential intention by the BIOS engineers. So for people on a similar build, I'd say just make sure to update your BIOS and forgo EXPO for the time being until we know things are stable and working as intended. The cause here is still speculative, however may be a good safety measure to take.
I did spend a few minutes looking for the pinout of AM5 to see if where it is damaged is associated with the memory controller. I couldn't easily find anything. I will now leave the investigation and speculation to the professionals 😉.
I did get the same hardware so wish me luck 😬. I am up and running with the updated BIOS version and I ain't touching any memory timings for the time being. Nothing else was damaged. As to why the same parts? There just isn't enough competition or aesthetically as pleasing hardware IMO. I still don't get in a world of RGB, manufacturers using non-neutral colors on products. I'm looking at you MSI with that gold MOBO... Remember when Corsair had a yellow logo on RGB products 🤮?
Anywho, thanks for the support folks! Let's see how this unfolds.
链接：https://www.reddit.com/r/Amd/comments/12tlk7s/comment/jhbfh8y/
大致总结：
OP的系统是默认，未进行任何CPU超频
OP的电脑内存是F5-6000J3038F16GX2-TZ5NR，在主板的QCL列表内，OP仅仅开启了EXPO
OP尝试过更新BIOS到1101, 但是更新后出现EXPO开启后不稳定和开机无法POST的情况，所以OP又把BIOS降级回了0922
OP 4.16拿到CPU后直至出事前一直在正常使用78X3D玩游戏，电脑也没有关机，运行游戏性能正常
78X3D暴毙出现于某天OP不玩游戏后，电脑空载的情况下，OP出门了，回来发现电脑暴毙，系统无法POST，QCode显示00
|
310112100042806
发表于 2023-4-23 21:10
