UPD: Looks like CPU actually does occasionally report 0 mHz clock, as Adrenaline logged similar behaviour... But how and why? And why EVERYTHING that is based on per-core basis reporting drops to zero?
Never noticed my overlay (RTSS) ever showing anything with 0's. And it updates 10 times as frequently (200ms vs 2s). And it continued working correctly even after HWINFO bugged out.
And yes, i tested at some point, issue with 0-roing minimal detected values still persisted even when i disabled...
Potential reason, graphical device being reset... Actual reason, no clue. Why CPU readings are ones that were affected - no idea. Why readings hadn't recovered, no idea. How to reproduce, no idea.
I already had long time issue of HWINFO randomly reading 0.000 V, 0 mHz and 0 degrees temps on...
Spoke too soon. . .
I have 2000 ms polling rate
Hmm... let me see result without snapshot polling for a chance
And it seems like only per-core polls are being messed up.
Without snapshot polling per-core VID is broken (it doesn't report correct values whatsoever). Temperature is also broken...
Maybe you're right then. ZenPT uses about 50 ms polling rate, so it definitely gives additional load.
Anyways thanks for insight. Will take this into account.
Hmm... Ok. It is just strange that even under load even 0 1 and 2 cores (Windows main, and top 1/2 performance) can get in gated state. But if it is possible. Wonder why i hadn't seen such thing via deep monitoring. Maybe because of additional real-time load?
No GPU drivers = No Adrenaline
No Afterburner
No Fan Control killed it via Task manager
Only HWINFO, basically.
And in first 5 minutes 3 cores reported 0.0 mHz effective clock and 0.0% utilisation.
I am not sure, maybe it is intended? but hard to believe that core 0 will ever have 0.0%...
I mean, of course i use HWINFO without it, as i said i don't run this tool constantly, it is taxing.
It actually takes processing power from Core 0 and reporting speed can get slower if there is high load on this core.
More than that. I also don't use this tool constantly explicitly because...
ZenPT Monitor, it is CPU focused tool. And specific to Zen based AMD CPU's
Granted some lines may be not named correctly, but most of them show what they must
Here is how it looks. Simple, but practical
UPD: Nope... Even without Afterburner and CPU monitoring in Adrenaline HWINFO still randomly reports 0.00 mHz. But only on effective clocks?
Can they actually be less than 0.1 mHz?
Nah, i only use them in sessions (they are too taxing, as their reporting speed about 10-20 refreshes per second). And when i use them i don't care much about HWINFO readings, as they are used for instantenous values. (And they really mess up HWINFO in terms of max clock readings. I saw up to...
Hmm... Sorry, but i will conntinue this thread despite several months passing.
Can it be something related to Windows 11 policies? Even though i do have Core Isolation disabled and no Hyper-V?
Because i have 2 deep CPU monitoring tool, one is with normal priority library, and second with Ring 0...
Finally, after total of 2 days with debug on (with some resets to reduce file size, i managed to catch 0.00 frequencies and 0.0 core count over night of foreground. No BS temp or voltage values though.
File is around 2.7 GB big, but i compressed it to around 77.5 MB
Link to debug file. Looked...