Perhaps Asrock has no fault other than setting more aggressive auto expo profiles at voltages adjacent to VSoc. Asrock has no way of contradicting the official AMD version
Please search, who is really interested and look at ALL Zen 5 that had burn marks where it corresponds on the AM5 diagram in the center side where the VDDIO pins are. No processor with visible burns or traces on the pins had them anywhere other than there, center side. I know, and the comments will start, because they know, because they heard what AMD officially says, tralala. Try to think a little logically, VSOC is generally under control, there are no more cases of 1.30V except extremely rare, but VDDDIO "flirts" freely in synchrony with VDDQ.
Everyone does as they please, but I say it's time to think outside the official statements, remember Intel 13-14 series when they initially blamed the mobo manufacturers. AMD is doing the same, but the problems will not stop, the cases will multiply, depending on each person's ram kit, some are more aggressive at VDD and VDDQ voltages, so automatically VDDIO is higher
Here is the AM5 diagram:
https://commons.wikimedia.org/wiki/File:Socket_AM5_pinmap.svg
Here my voltages,no error in over 6 months: