HWiNFO 64 (v6.27-4185) shows high voltages which do not match the voltages shows by Ryzen Master?

[Itay]

Hello!
I have a Ryzen 5 3600x and HWiNFO 64 (v6.27-4185) seems to report way higher voltages than those that Ryzen Master reports.
For instance, in the picture below you can see that Ryzen Master reports 1.209 peak and 1.0363 average while HWiNFO 64 (CPU Core Voltage SVI2 TFN) reports 1.456V current and 1.415 average which is extremely high and I don't know how safe it is for daily use.

Is that a bug? Which is the correct voltage?

https://imgur.com/fr5RvdK

 

[Martin]

The SVI2 TFN voltage shown by HWiNFO is the true voltage provided by the VRM to CPU.
What RM is showing is quite odd and frankly no one knows what exactly this represents. Some think this is voltage averaged by core load factor, so not the real voltage provided to CPU.

 

[Itay]

Thanks Martin.
So does that mean something is wrong? I don't think my CPU should almost always be at those voltages..

 

[Martin]

No, this doesn't mean anything is wrong.. Depends on type of load put on the CPU. Try with different loads - idle, 1C load with boost, all core load..

 

[stratego]

The voltage reported by Ryzen Master probably is Core VID

 

[Globespy]

I'm seeing exactly the same thing as the OP.
New 5800x on Gigabyte Aorus Master X570 (BIOS F34)
Win 11
Latest build of HWinfo64 running
Using PBO with negative -10 all core.
Turned off 'Cool n Quiet' and C-States, and haven't even loaded XMP profile yet.
I am seeing 'CPU Core Voltage (SVI2 TFN)' constantly sitting between 1.409v and 1.500v just sitting on desktop typing this email - if these readings are correct it will destroy my chip.

I turned the 'Cool n Quiet' and C-States back to enabled/auto and I see the 'CPU Core Voltage (SVI2 TFN)' sometimes goes down to 0.90v, but spends most of it's time in the mid/high 1.4v range.
When running a heavy stress test (running RealBench right now), I see this value dropping to between 1.3-1.368v on average.
CPU temps are solid in the 75-78C range entire time during the RealBench stress test.

I'm just concerned the 'CPU Core Voltage (SVI2 TFN)' idle voltages are dangerous to the life of the chip?
Although I just read a post on Reddit that seemed to suggest AMD have said this is 'normal':

https://www.reddit.com/r/Amd/comments/jxa4gz/_/gd4q79u

System seems stable but would appreciate some feedback - hopefully there's more information about Zen 3 since it's been out a while since this post was started.

 

[Martin]

No worries, that won't destroy your chip. The SVI2 TFN voltage reported in HWiNFO is correct as there's only a single voltage supplied to the entire CPU core domain.
What Ryzen Master shows is some artificial average voltage weighted with core activity and I don't believe this is something one can measure anywhere in the CPU.

You might also try to activate the "Snapshot CPU Polling" option in HWiNFO to get more accurate idle data.

 

[Globespy]

Thanks Martin!
I have 'Snapshot CPU Polling' enabled and doesn't make much difference.
I should see what this value looks like at BIOS defaults to see if maybe it's a problem with BIOS F34?

Good to know my CPU is ok for everyday use, wish there was more information about this as many are coming to AMD from Intel (like me) for the first time in years and these kinds of CPU core voltages are literally guaranteed degradation on modern Intel CPU's.

 

[Globespy]

Update:
Loading optimized (stock) BIOS defaults then booting into Windows see's almost identical behavior with this sensor.
It sometimes dips into the high 1.3v range but mostly in the 1.4v range.
So this must be by design.

 

[Martin]

It's still safe and the CPU has several mechanisms to prevent damage or reliability degradation (unless you override some settings like PBO scalar).