z790 vcore and VR out discrepancy

sta02

New Member
Anyone with Asus z790 noticed that there's an issue with reported vcore (die sense, z790 Hero) and EC VRM VOut reading? Seen it now on HWinfo 800, 802 and 804-5470. I believe they are more or less measuring the same thing, just at different places, so the numbers should be quite close (which they generally are)? However, EC sensor VOUT will regularly record a max value spike of around 1.53v, while the Nuvoton 6798D Vcore will record around 1.28-1.3v, which is what I'd expect. This happens even with auto cpu voltage cap and IA VR limit capped to 1.42v, CPU (14900KS) running at 5.7GHz. Can happen at any time, doing anything, just not idling. sometimes takes minutes, sometimes takes hours. Can be playing new games, old games, browsing, benchmarks.
Saw this on an Asus z790-F too, so it's probably not a faulty motherboard sensor? z790 hero is running most recent BIOS (2402, the one with Intel's etvb fixes) It still happens even if etvb and all the artificial boosting beyond the rated 5.7GHz max turbo freq alogrithms are disabled. Disabling c-states makes no difference. All CPU power related settings running at default Intel approved. Not running any other monitoring software. I don't have a debug file sorry.
 
As you said - those sensors measure the voltage at different places. VR VOUT is measured internally by the VRM, while Vcore is measure on the CPU side and there's difference due to Vdroop.
 
Thanks for your reply. The Vdroop I can see- VR VOUT consistently tracks a couple of hundredths higher than whatever Vcore shows (It's on LLC4 if I didn't mention it earlier). But then VR VOUT will hit 1.53v, with no real change on Vcore at all (which will still be around 1.28). This can happen while watching youtube, so not exactly high stress stuff. Is that still expected vdroop behaviour? Even with IA VR limit capped below 1.53v? Which number would be a better representation of the real voltage?
 
Back
Top