Strange Vcore Readings

[Calligo]

Hey!
I was building a new watercooled workstation (i7 8700k, Asrock z370 Extreme 4, G.Skill DDR4-3000..) and I tried to overclock my CPU a bit in the last few days, but I ran into a problem with HWInfo (- at least I hope its a HWinfo problem).

I get strange VCore readings although I'm on "fixed" Voltage (1.28V) via UEFI. HWInfo shows sometimes spikes of 3.1v, which is pretty much impossible. Furthermore on "auto" voltage my UEFI shows me about 1.168V, but in Windows I get spikes up to 1.44V after booting the system (idle).  :s

Tbh I didnt see those spikes happening irl, as this happens pretty rarely, but I run HWInfo in background the whole time, so I get the alarm windows when hitting >1.4V (thats why I cant compare it to CPU-Z readings - I would have to check CPU-Z 24/7 to maybe get a chance to catch it).

Well, I'm a bit worried now, if this is only a reading error of HWinfo or if something is very wrong with my board or even my CPU.

FYI: I also tried the new beta as well as the newest release to see if it fixes this issue, but it happens in both versions.

Is anyone aware of this? Does anyone have similar problems or is it just me?

 

[Martin]

3.1V is definitely a wrong readout. Are you perhaps using some other monitoring tools in parallel with HWiNFO? That might be a reason of occasional wrong values.
You might try to check with some other tool that shows min/max values like HWMonitor for example.

 

[Calligo]

No, I dont use any other tools. Those readings are happening in idle, when watching Netflix or doing similar tasks. The problem is.. Now I have to question every reading I get from HWinfo. Is the voltage reading on "auto" voltage correct? Does it show the real SA and VCCIO voltages or not, since I get alarms from those too. My UEFI shows me about 1.168v, but in HWInfo there are spikes up to 1.44v after booting into windows, are those real? Its pretty confusing and potentially dangerous. And now I also found a lot of other threads here about this issue, dating back to 2016. Is there an old HWInfo release out there without this issue? Since I have brandnew hardware and because I found lots of other threads about this, I dont think its a hardware problem or something, but of course Im not 100% sure.

I will check for an alternative to see if it happens there too.
Thanks for your help.

 

[Martin]

I don't think that the problem is in HWiNFO, it just shows values read straight from the hardware sensor, which is a quite simple operation.
If other tools show the same spikes, I'd say those values might be real, or there's some problem with sensors on the board.

 

[Calligo]

The thing is... I have now read dozens of threads from reddit over tomshw to asus and a lot of people seem to get wrong readings from almost all utility programs across the board. So I reached out to resellers, Asrock and some OC-forums, but nobody knows anything about it (Tbh I would be really surprised of nobody blew up their CPU by now because of this, but I found nothing but complains).
Some expect the power saving modes etc. to be the issue, but I cant reproduce it.

Ill check with HWMonitor and others, but it might take a while since I cant force those wrong readings.