[SOLVED] FX6300 236 Deg C High Temp reported in HW Monitor

[runnermatt]

Hi everyone, this is my first post here. I've googled high core temp XXX temp XXX or something similiar and I either get returns that say the max sustained core temp for AMD FX processors is 64 Deg C or other posts recommending replacing the thermal paste. This isn't exactly something I monitor regularly so I can't say if this is 100% new or has been going on for a while. I've been a electronics tech for about 20 years, but PCs are not my area of expertise. Hardware list at bottom of this post.
Anyway, the potential problem. On the 11th of October I opened HW Monitor and then started up Star Citizen and went into the PTU to see what my high temps and all looked like while in the game. I took screen shots to after to keep track. HW Monitor showed the below table numbers after that first session.
All of these numbers have been pretty much in line with the above since then. The only exceptions have been the high and low temperatures reported by the CPU. After a session of SC a couple nights ago the CPU "Cores" Max was reported as 161 Deg C.
Fallout 4 doesn't strain my PC any and I tried that last night and the reported high was 193 Deg C. Today I went a bought new thermal paste from BestBuy and redid the thermal paste. And had HW monitor open while scrolling reddit. Looked at HW monitor and saw a MAX over 225 Deg C and a Min of 0 Deg C. I closed HW Monitor and reopened it to clear the values and have been watching it while typing this up. At some point it reported a MAX of 224 Deg C.
My version of HW Monitor is 1.44.0 Driver version 150.

Soo, is my CPU dying or am I seeing a glitch in HW Monitor??? I would expect if these were real numbers the PC would power off. But as I said, PC are not my tech skillset.


ASUSTeK Computer Inc, M5A99FX
+Voltages
-Temperatures
CPU Value: 31 Deg C Min: 30 Deg C Max: 43 Deg C
Mainboard Value: 26 Deg C Min: 26 Deg C Max: 29 Deg C
+Fans
+Fans PWM
+Utilization
System Memory Value: 10% Min: 10% Max: 58%
AMD FX-6300
-Temperatures
Package Value: 10 Deg C Min: 9 Deg C Max: 29 Deg C
Cores Value: 10 Deg C Min: 9 Deg C Max: 29 Deg C
-Utilization (Low was 000% on each core and overall, High was 100% overall and on each core.)
-Clocks (Low was 1517MHz on each core, High was 3810MHz on each core)

All of these number have been pretty much in line with the above since then. The only exceptions have been the high and low temperatures reported by the CPU

Motherboard: ASUS M5 A99FX R2.0 PRO
CPU: FX-6300 (mild overclock applied by motherboard)
RAM: 4 x 8GB Corsair CMY16GX3M2A1600C9 (DHCP on)
GPU: MSI R9 290 4GB Gaming Edition
PS: 800W Phanteks unit
SSD: Seagate FireCuda (replaced HDD that failed 3 weeks ago)

 

[CountMike]

Hi everyone, this is my first post here. I've googled high core temp XXX temp XXX or something similiar and I either get returns that say the max sustained core temp for AMD FX processors is 64 Deg C or other posts recommending replacing the thermal paste. This isn't exactly something I monitor regularly so I can't say if this is 100% new or has been going on for a while. I've been a electronics tech for about 20 years, but PCs are not my area of expertise. Hardware list at bottom of this post.
Anyway, the potential problem. On the 11th of October I opened HW Monitor and then started up Star Citizen and went into the PTU to see what my high temps and all looked like while in the game. I took screen shots to after to keep track. HW Monitor showed the below table numbers after that first session.
All of these numbers have been pretty much in line with the above since then. The only exceptions have been the high and low temperatures reported by the CPU. After a session of SC a couple nights ago the CPU "Cores" Max was reported as 161 Deg C.
Fallout 4 doesn't strain my PC any and I tried that last night and the reported high was 193 Deg C. Today I went a bought new thermal paste from BestBuy and redid the thermal paste. And had HW monitor open while scrolling reddit. Looked at HW monitor and saw a MAX over 225 Deg C and a Min of 0 Deg C. I closed HW Monitor and reopened it to clear the values and have been watching it while typing this up. At some point it reported a MAX of 224 Deg C.
My version of HW Monitor is 1.44.0 Driver version 150.

Soo, is my CPU dying or am I seeing a glitch in HW Monitor??? I would expect if these were real numbers the PC would power off. But as I said, PC are not my tech skillset.


ASUSTeK Computer Inc, M5A99FX
+Voltages
-Temperatures
CPU Value: 31 Deg C Min: 30 Deg C Max: 43 Deg C
Mainboard Value: 26 Deg C Min: 26 Deg C Max: 29 Deg C
+Fans
+Fans PWM
+Utilization
System Memory Value: 10% Min: 10% Max: 58%
AMD FX-6300
-Temperatures
Package Value: 10 Deg C Min: 9 Deg C Max: 29 Deg C
Cores Value: 10 Deg C Min: 9 Deg C Max: 29 Deg C
-Utilization (Low was 000% on each core and overall, High was 100% overall and on each core.)
-Clocks (Low was 1517MHz on each core, High was 3810MHz on each core)

All of these number have been pretty much in line with the above since then. The only exceptions have been the high and low temperatures reported by the CPU

Motherboard: ASUS M5 A99FX R2.0 PRO
CPU: FX-6300 (mild overclock applied by motherboard)
RAM: 4 x 8GB Corsair CMY16GX3M2A1600C9 (DHCP on)
GPU: MSI R9 290 4GB Gaming Edition
PS: 800W Phanteks unit
SSD: Seagate FireCuda (replaced HDD that failed 3 weeks ago)

CPU-ID HW Monitor and AMD never plays well together. For AMD FX processors and system use AMD Overdrive https://www.guru3d.com/files-details/amd-overdrive-download.html

 

[drea.drechsler]

Hi everyone, this is my first post here. I've googled high core temp XXX temp XXX or something similiar and I either get returns that say the max sustained core temp for AMD FX processors is 64 Deg C or other posts recommending replacing the thermal paste. This isn't exactly something I monitor regularly so I can't say if this is 100% new or has been going on for a while. I've been a electronics tech for about 20 years, but PCs are not my area of expertise. Hardware list at bottom of this post.
....

You might like to know...although it doesn't help much...but with FX processors in particular AMD pretty much messed up temperature reporting.

Here's a short write-up and discussion about the issues.

IMPORTANT - CPU Core Temperature Measuring - Facts & Fictions

Intel CPUs: Later Intel CPUs feature a Digital Thermal Sensor (DTS) which measures the temperature for each core. Since many users are confused about certain values measured, there is a clarification needed. First, the value read from the DTS is not an absolute values in degrees Celsius, but...

www.hwinfo.com

 

[DSzymborski]

Story told simply: AMD temperature reporting of this era was garbage and if it was actually 236 degrees, it would no longer be a CPU but scrap metal entombed inside other bits of new scrap metal.

 

[runnermatt]

CPU-ID HW Monitor and AMD never plays well together. For AMD FX processors and system use AMD Overdrive https://www.guru3d.com/files-details/amd-overdrive-download.html

Thanks for the reply. I found AMD Overdrive last night after scanning some other posts for FX series processors. Downloaded and installed it last night. Found the Logging function in Overdrive this morning. Going to run a log now concurrently with HW Monitor and compare the two so I can hopefully find I have nothing to worry about.

 

[runnermatt]

You might like to know...although it doesn't help much...but with FX processors in particular AMD pretty much messed up temperature reporting.

Here's a short write-up and discussion about the issues.

IMPORTANT - CPU Core Temperature Measuring - Facts & Fictions

Intel CPUs: Later Intel CPUs feature a Digital Thermal Sensor (DTS) which measures the temperature for each core. Since many users are confused about certain values measured, there is a clarification needed. First, the value read from the DTS is not an absolute values in degrees Celsius, but...

www.hwinfo.com

Thanks for the reply. I'm read up now and things are clearer and more confusing at the same time now. Seems like some engineer had "a good idea" and there wasn't anyone that pushed back or maybe wanted to see the result. I don't have a problem with engineers wanting to see the result, usually.

From the reading it seems that the sensors are:
one). Setup to report the difference from temperature max. And,
two). Accurate closer to temperature max and more inaccurate the further the reading is from temperature max.

So maybe, it is running cool enough that the reading is highly inaccurate and what is reported as the Core temp and HW Monitor is interpreting it as a unrealistically high temperature.

I figured HW Monitor was just displaying what the hardware was reporting and didn't think it could be wrong. I also know solder melts at about 300 Deg F(148 Deg C) so I expected if the CPU was actually getting that hot it was either for so short a time period that it wasn't having instant catastrophic damage OR the CPU sensor was going bad OR the CPU wasn't getting that hot.

Glad to find that it is likely just software/hardware misreporting.

Thanks for the link.

 

[runnermatt]

Story told simply: AMD temperature reporting of this era was garbage and if it was actually 236 degrees, it would no longer be a CPU but scrap metal entombed inside other bits of new scrap metal.

Thanks for the reply.

Yeah, I know solder melts at around 300 Deg F (148 Deg C). So I couldn't foresee anything electronic being able to withstand those temperatures. I thought maybe if it happened super fast and cooled back down, just maybe. The fact that the reported highs were getting higher made me kinda suspect that maybe it was going higher because something was degrading.

From the other two posts and links and doing some more reading it appears to likely just be a miscommunication from the hardware and HW Monitor mistranslating it.

 

[drea.drechsler]

Thanks for the reply. I'm read up now and things are clearer and more confusing at the same time now. Seems like some engineer had "a good idea" and there wasn't anyone that pushed back or maybe wanted to see the result....

I'm really not that certain things are getting any better...especially if you're OCD about temperature monitoring. The Ryzen CPU's also have very strange temperature readings and it's also not considered a 'true temperature' but more an indication of the thermal state of the CPU die. As geometry shrinks CPU dies are getting smaller and smaller right along with the transistor junctions on them. That results in something they're calling a 'power density' problem, where too much power is being dissipated in such a small volume with inadequate surface area to transfer the heat out. Leads to weird things like extreme spiking temps in the tiny region of the die close to a temp sensor when a core boosts even though the bigger CPU is actually quite cool overall.

I've read an article that as geometry shrinks from 7nm (Zen 2 and 3) to 5nm (I think Zen 4) AMD's re-thinking the whole temp reporting thing even again. So hang loose on new CPU's as it gets even stranger.