What's the difference between "CPU Core Voltage" and "Vcore" in HWInfo64?

[trents]

What's the difference between "CPU Core Voltage" and "Vcore" in HWInfo64?

HWInfo64 shows a line under the CPU readouts section called "CPU Core Voltage (SV12 TFN)" that corresponds to the value I enter in bios for "Override CPU Core Voltage. Under load it drops a little bit as you can see at the LLC level I am currently using.

HWInfo64 shows a line under the motherboard readouts section called "Vcore" that actually rises significantly under load and, as you can see, exceeds 1.48 volts.

What is the difference between these two parameters. What is each actually measuring? Is "Vcore" here measuring VRM current?

I have also included a pic of CPU-z for comparison's sake.

Do you see any unsafe values here?

 

[EarthDog]

I do not know the answer to your question...but I did find this on a quick google search:

Core Voltage under the CPU section is a sensor located on the CPU itself and Vcore under the motherboard section is a sensor located on the board most likely in or near the VRM.

So, it may have something to do with the VRM, but if it was measuring current, the value would be in Amps, not volts. So perhaps it is just the voltage post VRM but before it hits the CPU?

If we can't answer it, Martin can!: https://www.hwinfo.com/forum/index.php

 

[Zerileous]

That would make sense, but on a smaller board I have a hard time imagining that much voltage lost in the short distance from the VRM to the CPU.

Does the vcore value hold that setting under certain conditions, or is it only a momentary spike?

For my board, I have sections dedicated to the VRM in HWiNFO64. The first section labeled with my motherboard "ASrock X470 Taichi" is also labeled "(Nuvoton NCT6779D)." The second ones are labeled "(IRF IR35204)." These can be read by moving the label tab over to the right. My CPU tab does not have this extra label. My "CPU Core Voltage" in the CPU section is labeled (SVI2 TFN). The "VR Out" reading in the IRF section which I believe to represent the Vcore VRM. I can't find much specific detail on the Nuvoton sensor, but it appears to be a common brand. I agree with EarthDog, this is a good question for Maritn, as it is bound to come up repeatedly. While by no means as dramatic, my system also displays a higher "Vcore" reading from the Nuvoton sensor. "VROUT" and "CPU Core Voltage" appear to be in agreement given a small amount of drop across the power plane. You can also note relatively high "Maximum" values for the "Vcore" sensor, although my CPU is using Precision Boost Overdrive, which is known to use high vcore for short periods of time. And of course a picture is worth a thousand words:

 

[trents]

Thanks for the suggestion ED. So I registered on Martin's forum and posted my questions.

Z, the high vcore readings are not momentary. They are sustained during heavy loading phases but drop off between heavy loading phases while running the Realbench stress test. Edited.

 

[trents]

Here's Martin's reply:

Martin Wrote:
CPU Core Voltage (SVI2 TFN) is the voltage value provided by the voltage regulator to the CPU (and the value the CPU uses for internal telemetry/power management).
This value is sensed directly by the VRM and in cases where it's possible to read VRM data, it should match the VOUT value from VRM.
Vcore is the value measured somewhere along the way via a third component (i.e. SIO like ITE or Nuvoton). This value is less precise.
This topic been discussed here already and especially on the overclock.net forum, where several members did measurements and explained in detail how this works and why there are differences (vdroop, etc.).

 

[Alaric]

One of my favorite motherboard features is the ability to measure critical voltages with a VOM. It has allowed me to "exceed" some voltage ceilings (as reported in software) and get that last smidgen of OC.

 

[ShrimpBrime]

The cpu core voltage is Processor request VID. The later voltage is actual output to cpu.
Looks like the just changed the description from VID to SV12 TFN.

 

[trents]

The cpu core voltage is Processor request VID. The later voltage is actual output to cpu.
Looks like the just changed the description from VID to SV12 TFN.

That seems at odds with what HWInfo64's author, Martin, says. If you look back in post #5 in his reply to this question, he says there, "CPU Core Voltage (SVI2 TFN) is the voltage value provided by the voltage regulator to the CPU."

 

[ShrimpBrime]

Suspicious. As the VID of that particular cpu has a VID of 1.35v while it should be 1.212v.

Must be overclocked or SenseMi tweaks cause a high VID reading.

I did read all the posts. Nobody mentioned processor VID....

 

[trents]

It is overclocked. I think that is reflected in the first line of my Sig. I will edit the Sig and make that clearer.

 

[ShrimpBrime]

All good. The readings look to be in spec then.
Sry if I misunderstood something here.