HWiNFO64 CPU Die (Average) Spike?

[llkhrdh]

Just to add more information about my setup


CPU: 7800X3D / MB: GIGABYTE B650 AORUS ELITE AX / RAM: Team T-CREATE EXPERT 32GB (2 x 16GB) DDR5 6000 (PC5 48000)
GPU: XFX RX 6700 XT / CPU cooler: Thermalright PA120 SE Extremely Double Assassin / PSU: EVGA 850 Bq, 80+ Bronze 850W

Bios Modification : only undervolted to -25
dual monitor: one via GPU and the other via motherboard(IGPU)
System: Win10 pro
Active programs when the spike happened : AMD GPU+CPU related softwares / HWinfo / Discord / notepad

 

[Martin]

i tried already, raw it was 900 megs compressed its 130ish

That's too big to attach or send directly. Can you upload it to some public sharing service or OneDrive, Google Drive?

 

[Deleted member 25033]

Just to add more information about my setup


CPU: 7800X3D / MB: GIGABYTE B650 AORUS ELITE AX / RAM: Team T-CREATE EXPERT 32GB (2 x 16GB) DDR5 6000 (PC5 48000)
GPU: XFX RX 6700 XT / CPU cooler: Thermalright PA120 SE Extremely Double Assassin / PSU: EVGA 850 Bq, 80+ Bronze 850W

Bios Modification : only undervolted to -25
dual monitor: one via GPU and the other via motherboard(IGPU)
System: Win10 pro
Active programs when the spike happened : AMD GPU+CPU related softwares / HWinfo / Discord / notepad

This is insanity you also have a peerless assassin se. Also a b650 elite ax. Up to now it looks like it isn't cpu exclusive nor motherboard but it is strange that so many people with this cooler are having the same issue including me. I wonder Is it possible to have cooler interference on the cpu temp sensor? Maybe its just that this cooler is popular or maybe there's an issue with thermalright but this is just speculating

My specs simplified are
CPU: Ryzen 7 7700x
GPU: rx7900xtx
MB: B650 Aorus Elite AX
CPU Cooler: Thermalright PA120 SE ARGB
PSU: Seasonic 1000w

Bios Modification : -30 undervolt all cores 85c limit and 85w ppt limit
Single monitor
System: Windows 11
Active programs: Firefox with 3 tabs open and playing fortnite landing at Hazy Hillside then entering the rebel hideout is when I seen the first spike. The cpu fan also ramped up during that time the cooler has done that before and only spiked to 87c

Edit I've found someone with the same spike issue on reddit:

https://www.reddit.com/r/AMDHelp/comments/143130m

 

[Martin]

Cooler - extremely unlikely.

 

[Deleted member 25033]

Cooler - extremely unlikely.

I agree that the chances of that would be really low but looking at the specs of everyone's pc its the only similar thing I can see. I'll keep testing and see what works for me because this is great software so I'll try to find out whats causing this bug

 

[Renkou]

I agree that the chances of that would be really low but looking at the specs of everyone's pc its the only similar thing I can see. I'll keep testing and see what works for me because this is great software so I'll try to find out whats causing this bug

I really doubt the cooler is the issue, It's only a piece of metal clamped to the chip. There's nothing plugged into the heatsink for it to even pull power unless you're counting fans but then that's the fans not the heatsink. I feel It's just a sensor reading error between the software and hardware. Like I mentioned earlier my vsoc is locked to 1.2v in the bios so therefor It should never ever move with it on manual mode. Other then that all readings are normal as seen below. Only other issue I see rarely, which is super minor to me and shows the software has an error with sensor reading, is sometimes when I'm afk and the pc idles and turns my monitors off, the software will read my lowest cpu temps as 0c when in reality it's lower 20sc on my peerless. On full idle i can get as low as the teens for cooling.

 

[Deleted member 25033]

I really doubt the cooler is the issue, It's only a piece of metal clamped to the chip. There's nothing plugged into the heatsink for it to even pull power unless you're counting fans but then that's the fans not the heatsink. I feel It's just a sensor reading error between the software and hardware. Like I mentioned earlier my vsoc is locked to 1.2v in the bios so therefor It should never ever move with it on manual mode. Other then that all readings are normal as seen below. Only other issue I see rarely, which is super minor to me and shows the software has an error with sensor reading, is sometimes when I'm afk and the pc idles and turns my monitors off, the software will read my lowest cpu temps as 0c when in reality it's lower 20sc on my peerless. On full idle i can get as low as the teens for cooling.

My bad yeah I meant there could be something up with the fans but your probably right with the temps mine got to surely it would have cooked itself

 

[Renkou]

That's too big to attach or send directly. Can you upload it to some public sharing service or OneDrive, Google Drive?

Here's the link

Renkou 12-26-23 7800x3d.zip

drive.google.com

 

[Martin]

Here's the link

Renkou 12-26-23 7800x3d.zip

drive.google.com

Thanks!
For the first time since this problem was reported I could analyze the raw data and I can say that these spikes are not a bug in HWiNFO.
I can clearly see that when this glitch happens, the CPU reports several parameters double of what it has read in the previous or subsequent poll cycle. Not just the temperature or VDDCR, VSOC, but several other parameters are reported this way.
This is clearly an erratum in CPU/SMU and AMD should look at that and fix it.

 

[Renkou]

Thanks!
For the first time since this problem was reported I could analyze the raw data and I can say that these spikes are not a bug in HWiNFO.
I can clearly see that when this glitch happens, the CPU reports several parameters double of what it has read in the previous or subsequent poll cycle. Not just the temperature or VDDCR, VSOC, but several other parameters are reported this way.
This is clearly an erratum in CPU/SMU and AMD should look at that and fix it.

Since no other program is catching this, what would be the proper route to take for this? just keep using the pc as is and see if anything drastic happens then rma? Is the cpu voltage actually spiking or are these all read errors on the manufacturers end with the CPU/SMU.

 

[Martin]

Those spikes (which seem to be exactly double of what it should be) are in fact invalid/misreporting. So no damage can happen because of this, it's just a misreporting bug that should be fixed by AMD. It should be possible to fix it a BIOS update once they locate the bug.

 

[Renkou]

Those spikes (which seem to be exactly double of what it should be) are in fact invalid/misreporting. So no damage can happen because of this, it's just a misreporting bug that should be fixed by AMD. It should be possible to fix it a BIOS update once they locate the bug.

Martin thank you soooo much and I'm happy I was able to help. If there is anything else I can do to assist let me know as ill lurk this post and stay active to see what happens with this issue. Hopefully AMD looks into this if they aren't already and apply a fix, if they aren't already.

 

[Deleted member 25033]

Thanks!
For the first time since this problem was reported I could analyze the raw data and I can say that these spikes are not a bug in HWiNFO.
I can clearly see that when this glitch happens, the CPU reports several parameters double of what it has read in the previous or subsequent poll cycle. Not just the temperature or VDDCR, VSOC, but several other parameters are reported this way.
This is clearly an erratum in CPU/SMU and AMD should look at that and fix it.

Great how do I delete my account?

 

[ThatKidRuss]

Should I be concerned about this reading or is this miscalculated by the application? My voltage double and my CPU DIE Average spike to 107 degrees.

Quick info about my specs:

7700X
B650E-E ROG STRIX
CORSAIR 16X2 6000MHZ
1000W CORSAIR HX PLATINUM
6800XT

 

[Steed]

Should I be concerned about this reading or is this miscalculated by the application? My voltage double and my CPU DIE Average spike to 107 degrees.

Quick info about my specs:

7700X
B650E-E ROG STRIX
CORSAIR 16X2 6000MHZ
1000W CORSAIR HX PLATINUM
6800XT

Based off of the image attached. Looking at the maximum numbers I believe this would fall under the misreporting issue.

 

[ThatKidRuss]

Based off of the image attached. Looking at the maximum numbers I believe this would fall under the misreporting issue.

This only happens when i play games for at least an hour. I was confused if this is true reading but i might just download HWmonitor to see if it can reproduce the same reading.

 

[Steed]

This only happens when i play games for at least an hour. I was confused if this is true reading but i might just download HWmonitor to see if it can reproduce the same reading.

So far from what I have read and experienced personally that falls in line with what seems to be happening, roughly for me at least about an hour in I see the doubling of the value.

 

[Martin]

I have submitted the data to AMD and they are analyzing it. But it's holidays and almost New Year, so we need to be patient...

 

[Renkou]

So far from what I have read and experienced personally that falls in line with what seems to be happening, roughly for me at least about an hour in I see the doubling of the value.

ya its at very random times also for me. Full load i get double values and on idle i get 0c temps on all cores lmao. its wonky for sure

 

[levy93]

Am I suffering the same problem too?