Processors

[cpurpe91]

I recently put together a system with an Ryzen 7 7700X in a MAG X670E TOMAHAWK WIFI, with DDR5 6000mhz c30 Corsair Dominator Titaniums using EXPO. This system runs flawlessly, with a hiccup. The hiccup is that so far two seperate programs have reported the same reading. At first I thought it was a bug concerning HWinfo64, and I saw many others are also dealing with it. Now I am not so sure what is going on. On several occasions I have dealt with these readings and I had largely ignored them. 

Here is a screenshot of OCCT being used to monitor my temps while I was game. I alt-tabbed to see this.

What should stand out to you is the 113C reading as well as the CPU VDDCR_VDD Voltage (SVI3 TFN), as well as the other two SVI3 TFN readings. Also note that thermal throttling was not triggered by the spike. 

The motherboard sensors are in the image below.

This is identical behavior to what I recently saw with HWinfo64.

Both of these images are cut from the same screenshot. Notice how the "spike" happened and no throttling was triggered.

This has happened a number of times to me so far and I am not certain whether I should stop using the machine or just ignore the readings of two different programs.

Here is the thread on HWinfo64 forum. 
https://hwinfo.com/forum/threads/hwinfo64-cpu-die-average-spike.8044/
https://hwinfo.com/forum/threads/hwinfo64-cpu-die-average-spike.8044/page-13

There have been numerous people reporting this issue to HWinfo forum team, but I figured maybe someone here may have an answer since it is now effecting two programs and not just HWinfo64.

I have posted my findings using OCCT to track temps while I was gaming on the HWinfo forum as well. 

[misterj]

cpurpe91, please compare these results with Ryzen Master (RM). Free HW monitors are notorious for reporting bad to really bad values. I do not use them and recommend that no one use them to make decisions. Enjoy, John.

[elstaci]

The reason probably why OCCT and HWinfo have the same or similar data is because I believe OCCT uses HWinfo as it monitoring engine.

From OCCT Website:

Try using another similar monitoring software like Open Hardware Monitor and  see if you still see those temperature spikes.

It might be a bug in that one specific program. I would believe that at 113c Desktop processor temperature the computer would have immediately shut down unless the spike was so fast that the CPU didn't have time to shut down or possibly the CPU was engineered to shut down after a certain amount of time at its critical shutdown temperature and the spikes are within that time limit. Just guessing since I am not an engineer.

[cpurpe91]

According to the HWinfo forum, AMD and HWinfo are investigating to find the cause of the spike. It's all in the HWinfo forum thread I linked. I recommend anyone else experiencing this issue to post on the HWinfo forum with their system specs in the linked thread in the post. I have seen it effecting Ryzen 5000 and Ryzen 7000 so far across Gigabyte, MSI, and Asus motherboards.

[elstaci]

So what those links is indicating is that AMD has started to verify or has started to investigate those high Temperature spikes on the CPUs?

If those high temperature spikes are also showing on other monitoring software similar to HWinfo then it is pretty much verified since various apps are showing the same thing. But if it is only showing on one App then I would need more information unless HWinfo has a different method of reading the CPU's or Motherboard Temperature sensors.

But if it can be verified I wonder if those spikes will damage the CPU in the long run since it only spikes for a very short time. Not really, in my opinion, enough time to damage the CPU but possibly the accumulated affects of the spikes might cause damage eventually to the processor.

[cpurpe91]

Reading the entire thread a HWinfo forum staffer has been in talks with AMD about it. AMD is working on it according to the staffer. I got really worked up and confused when OCCT did the same thing but now knowing it is powered by HWinfo, it makes more sense. Any news on the issue I will post here.