For the same system in a previous article, AMD has been very helpful, and they wanted to see if having a newer BIOS on the motherboard with a newer version of AGESA would help things out any. Seems a reasonable request, as the BIOS on the motherboard was a year old, so I flashed a new BIOS and tried it again.
Trouble in paradise manifests around 11m:50s, and AMD has sent me a short RMA form to fill out and return.
Updated information from this time around:
Motherboard BIOS Version:
- Before flashing: E7A34AMS.120 03/2017
- After flashing to newer version: E7A34AMS.1C0 01/2018
New Bios Screenshots:
BIOS Screenshot - Shows updated BIOS version.
BIOS Screenshot - Shows voltage readings under new BIOS. (Spoiler alert: they haven't changed.)
- Ambient (room) temp about 22 degrees C.
- No appreciable load, around 30 - 35 degrees C according to k10temp
- Under load, peak in the low 50s degrees C per k10temp
- Temperature at failure point, 52.4 degrees C according to k10temp
New, Improved Information Video:
- Slideshow, sensors for k10temp, top, tail -f /var/log/kern/log
- https://github.com/Oxalin/ryzen-test (very mild stress test)
- Camera monitoring inside of PC (showing cooling live as it happens)
- At failure time (11m:50s), evidence appears in ryzen-test and tail kern.log windows.
- Video online: https://youtu.be/26x56Ft8EZ8
- Video download: The_AMD_Ryzen_Linux_Testing_Part_2.mkv [138MB]