Certain mainboard temperatures misreported



since a couple of days, maybe for two weeks now, some mainboard temperatures are misreported as being negative. Only certain temperatures are affected, e.g. mainboard/CPU and, curiously, only one of two MOS values.

I observed this bug on Windows 11 first and, after a new installation, on Windows 10 right now.
HWiNFO-Version is the latest one (7.22-4731), just updated before creating the debug file.
There is a service running in the background utilizing LibreHardwareMonitor, but stopping that and restarting HWiNFO changes nothing.
Normally the values correct themselves after some time, so I'm not really sure what's/who's at fault here.

Mainboard is a ASRock Z370 Taichi.

Debugging file is attached. I had "Accept negative temperatures" disabled initially and enabled it some time into the debugging.


  • HWiNFO64.zip
    672.6 KB · Views: 1
Try to reboot the system and run HWiNFO without starting LibreHardwareMonitor.
As expected the values are normal again, but that's always the case after a reboot.

What I noticed is that the LHW service seems to read at least one of the values in question (the MOS one) properly.

I'll leave the system at this state (LHW-service stopped) for a couple of hours and see if that error appears again.
Do the values in HWiNFO start to show invalid/negative after you run LibreHardwareMonitor?
Are you perhaps running some other monitoring tool? My assumption is that something in the system is reconfiguring the SIO hardware monitor.
Nothing that would really stand out, only software I've been using for quite some time without any problems regarding HWiNFO.
MSI Afterburner, RTSS, ThrottleStop, VMware, some gaming clients