HWiNFO64 CPU Die (Average) Spike?

[JasonVoorhees]

If anyone is using FanControl or LibreHardwareMonitor along with HWiNFO, try without them as they can interfere with HWiNFO.

Does any software that measures temps and voltages cause a conflict with Hwinfo or is it just those two? I wrote in a previous post of mine that when Hwinfo gave me a warning for exceeding voltages and temperatures , AMD's Adrenaline software registered those spikes too but I havent seen those spikes yet when solely using the metrics overlay(Adrenaline) and graphs for both CPU Temps and Voltage to monitor it on the go. Is it even possible that Hwinfo, or whatever the culprit for this is, could cause this behaviour to appear on a different measuring software?

 

[Martin]

Not any software, only some that use certain interfaces and lack interlocking mechanisms.

 

[JasonVoorhees]

Not any software, only some that use certain interfaces and lack interlocking mechanisms.

So given the fact that it happened with/on Adrenaline, im assuming it is one of them? Also, i dont know if you already knew this , but i just wanted to inform you that its not always doubling the readout, from the numbers i could observe it also sometimes adds 16,67%, 25% or 67%.

 

[PewPew]

Since v8.00 didn't have any changes for this issue, the difference was probably something else you did.
Yes, please go ahead with v8.01-5425.

Like suspected, this v8.01-5425 also gives me correct values.

So I think we're onto something here, because the only thing I changed in my setup that made the values correct even in v8.00 was to correct the error that showed up in event viewer with ID 7000 failed to load AMDRyzenMasterDriverV22 and having both AMDRyzenMasterDriverV20 and AMDRyzenMasterDriverV22 entries in registry.

Removal with AMDCleanupUtility in safemode, and clean installing the chiptset driver after it, corrected my readings with HWiNFO I suspect.

Again, for people having this problem please check HKEY_LOCAL_MACHINE\SYSTEM\ControlSet001\Services in regedit and scroll down a bit to check if you have 1, or 2 diferent version entries in there to check if it is the cause.

Just to be sure I'll use next few days the 7.72 version that I first reported here giving me double readings, if it gives me correct values then we'll know it is system related, and not HWiNFO related.

 

[PewPew]

Unfortunately it is not the case of being system related and not HWiNFO related, because in the old version 7.72 the double readings came back. Also the Average Active Core Count goes above the CPU 6 cores.



So I don't know but something is diferent for me between the v7.72 and the v8.00, because v8.00 and newer doesn't show me these readings, even using other metric tools as AF or AMD OSD.

 

[ToyTen]

Thank you for this 802 vesion, it looks very good so far. With FanControl active.

 

[amoosing]

Same problem here. Random spikes to 150ish C. Usually while playing (started Cyberpunk recently)
7950X3D on AsRock B650E PG-ITX WiFi. FanControl present. Just before finishing reading the thread I updated 8.00->8.02.
Will provide more info if it happens again after update.

 

[Zach]

Same problem here. Random spikes to 150ish C. Usually while playing (started Cyberpunk recently)
7950X3D on AsRock B650E PG-ITX WiFi. FanControl present. Just before finishing reading the thread I updated 8.00->8.02.
Will provide more info if it happens again after update.

You have a lot more double values than that 150C temp.
I can identify at least 9 more

 

[Martin]

Same problem here. Random spikes to 150ish C. Usually while playing (started Cyberpunk recently)
7950X3D on AsRock B650E PG-ITX WiFi. FanControl present. Just before finishing reading the thread I updated 8.00->8.02.
Will provide more info if it happens again after update.

HWiNFO64 CPU Die (Average) Spike?

Yes, AFAIK FanControl is using LibreHardwareMonitor.

www.hwinfo.com

 

[ToyTen]

Feedback of the sit here: no more msissreadings, none. v802 and FanControl active.

It is the second day without "issue"

Regards

 

[amoosing]

Yes, after upgrade yesterday, no more outliers.

 

[JasonVoorhees]

Still have faulty readings( Or are they?) after updating. However i noticed its by a much smaller margin(around 7%) and i have so far noticed this behaviour with vsoc, frequency limit,and uclk values. It happens everytime i play or do another task that is cpu intensive. The only monitoring software besides hwinfo is adrenaline, but i turned off all metric loggin and metrics overlay in general. It still does monitor some temps and cpu frequency in the background though so i dont know if that would have changed anything.

 

[JasonVoorhees]

Yes, after upgrade yesterday, no more outliers.

UCLK, FCLK, VSOC and Frequency limit are still multiplied by roundabout 7% for you. It also happens for me.

 

[Zach]

Still have faulty readings( Or are they?) after updating. However i noticed its by a much smaller margin(around 7%) and i have so far noticed this behaviour with vsoc, frequency limit,and uclk values. It happens everytime i play or do another task that is cpu intensive. The only monitoring software besides hwinfo is adrenaline, but i turned off all metric loggin and metrics overlay in general. It still does monitor some temps and cpu frequency in the background though so i dont know if that would have changed anything.

Can we see a whole window of HWiNFO sensors please? That 7% seems a little weird.

 

[JasonVoorhees]

Can we see a whole window of HWiNFO sensors please? That 7% seems a little weird.

Hmm, thats weird. This time i left Hwinfo open all day and when i came back to check on it the readouts were doubled again instead of the 7% that i saw yesterday. This time i didnt play or use anything cpu intensive, so thats a bit strange. But you can see the 7% increase in the image "amoosing" attached to his message aswell.

Edit: I forgot to turn off adrenaline metrics logging and overlay so maybe that is what caused the x2 readouts. Will turn it off and keep hwinfo running in the background one more time and report back.

 

[JasonVoorhees]

Ok, so now that i turned off adrenaline metrics, it happened again after some time. I was browsing the web and when i came back to check, i noticed that some readouts were 6,65% higher than what they usually are. I've marked them, but there might be some i havent noticed yet. (SVI TFN3 was marked by mistake)

Can we see a whole window of HWiNFO sensors please? That 7% seems a little weird.

 

[ToyTen]

9 days without any missreadings here.

Thank you

 

[Zach]

Ok, so now that i turned off adrenaline metrics, it happened again after some time. I was browsing the web and when i came back to check, i noticed that some readouts were 6,65% higher than what they usually are. I've marked them, but there might be some i havent noticed yet. (SVI TFN3 was marked by mistake)

Did a test gaming session... I saw the same.
At least for my FCLK/MCLK/UCLK max speed readings, all was peaked at +6.67% from actual settings (1800MHz >> 1920MHz)
HWiNFO v8.02-5440

 

[JasonVoorhees]

Did a test gaming session... I saw the same.
At least for my FCLK/MCLK/UCLK max speed readings, all was peaked at +6.67% from actual settings (1800MHz >> 1920MHz)
HWiNFO v8.02-5440

Hmm, okay. The bizarre thing, atleast to my unschooled eye, is that these readouts seem all to be memory related.

 

[Martin]

Folks, we have a new build to try: www.hwinfo.com/beta/hwi64_803_5451.zip
Please let me know the results.