Truenas shows almost 90°C, cpu fanspeed stays low...

[jagdtigger]

Hi all.

So i just noticed that the rating in the GUI seems a bit off. It shows ~54 °C as idle temp and ~90 °C at approx half load.
CPU: 1700X
Cooler: BeQuiet Silent Loop 120 AIO

I checked and the BIOS shows ~35 °C for idle temp. This and the fact that the aio could handle almost triple of the tdp of the 1700x leads me to believe this is some sort of bug.

Whats the community's opinion this?

 

[Kris Moore]

That does sound like a bug of some sort. If you open a shell prompt and run:

# sysctl -a | grep temperature

What do you get back? Same bad temps or something different?

 

[ornias]

Your motherboard is a little more relevant than your CPU or cooler tbh... (considering the discrepancy between the two temperature readings)

 

[jagdtigger]

That does sound like a bug of some sort. If you open a shell prompt and run:

# sysctl -a | grep temperature

What do you get back? Same bad temps or something different?

root@zenifer[~]# sysctl -a | grep temperature
dev.cpu.15.temperature: 87.5C
dev.cpu.14.temperature: 87.5C
dev.cpu.13.temperature: 87.5C
dev.cpu.12.temperature: 87.5C
dev.cpu.11.temperature: 87.5C
dev.cpu.10.temperature: 87.5C
dev.cpu.9.temperature: 87.5C
dev.cpu.8.temperature: 87.5C
dev.cpu.7.temperature: 87.5C
dev.cpu.6.temperature: 87.5C
dev.cpu.5.temperature: 87.5C
dev.cpu.4.temperature: 87.5C
dev.cpu.3.temperature: 87.5C
dev.cpu.2.temperature: 87.5C
dev.cpu.1.temperature: 87.5C
dev.cpu.0.temperature: 87.5C

Your motherboard is a little more relevant than your CPU or cooler tbh... (considering the discrepancy between the two temperature readings)

Code:

Handle 0x0001, DMI type 1, 27 bytes
System Information
    Manufacturer: Micro-Star International Co., Ltd.
    Product Name: MS-7A33
    Version: 2.0
    Serial Number: To be filled by O.E.M.
    UUID: 00000000-0000-0000-0000-4ccc6afb4bb2
    Wake-up Type: Power Switch
    SKU Number: To be filled by O.E.M.
    Family: To be filled by O.E.M.

Handle 0x0002, DMI type 2, 15 bytes
Base Board Information
    Manufacturer: Micro-Star International Co., Ltd.
    Product Name: X370 GAMING PRO (MS-7A33)
    Version: 2.0
    Serial Number: <redacted>
    Asset Tag: To be filled by O.E.M.
    Features:
        Board is a hosting board
        Board is replaceable
    Location In Chassis: To be filled by O.E.M.
    Chassis Handle: 0x0003
    Type: Motherboard
    Contained Object Handles: 0

 

[Revilio]

Have you ever tried to stress your CPU in other OS (i.e. prime95 on Windows)? If yes, what is the temp?

 

[jagdtigger]

Back when i assembled it i used windows to test it but that was 2 or 3 years ago so IDK what the temps looked like.... Ill try it with win2go after handbrake finished with its queue.....

 

[Bikerchris]

Only a guess but I wonder if it's an AMD CPU issue?

 

[jagdtigger]

Dont know, in theory if the cpu was really 80+ °C the fans would spin at full rpm non-stop. Which isnt the case, handbrake still has 4 tasks left ATM....

 

[jagdtigger]

Okay, handbrake done, running p95 now. But if i had to guess bsd reads out the cpu tctl instead of tcdie......

/EDIT
After ~10 minutes it seems it will cap out at 80.3 tcdie, tctl is 100.3...

/EDIT2
Final results after 20 or so minutes:
tcdie 82.3 max
tctl 102.3

 

[jenksdrummer]

May want to pull the heatsink/pump/etc, and put on new thermal goop after cleaning the old stuff off. I have seen some brands harden up or there's an air bubble, and causes issues.

To that point, get an in-OS temp monitoring software; free stuff (openhardwaremonitor or CPUID HWMonitor; no need to bother with any "paid" versions) - run them, and see what it shows idle, under load, and so on.

 

[jagdtigger]

Dug around and finally found the answer:

CPU (Tctl): This is the T_control temperature available on AMD CPUs only. On several generations before Zen (Ryzen), this is not a reliable representation of the temperature. On AMD Zen series this is the temperature used to control cooling and is a fixed offset from the real CPU temperature. Offset is used mostly on X-series and some Threadripper CPUs; in such case two values are shown: Tctl and Tdie. If no offset is used, then only a single value is shown as Tctl/Tdie, which equals the real temperature.

-CPU (Tdie): This value is shown in case the CPU uses an offset from Tctl and represents the real temperature (Tdie = Tctl - Tctl_offset).

What are the difference between these temps ?

Hi In HWinfo app it display CPU temperature in 3 values: Tctl/Tfie, average, Tdie . What are the difference between them ?

linustechtips.com

So the monitoring SW actually reads out the wrong value.... That 80 °C actually is pretty good for running basically a power virus.

 

[marshalleq]

I've just come here because I'm trying out TrueNAS SCALE and found it's misrepresenting the CPU temp - 70 degrees C at idle / 0% load. I'm already aware of the Tdie etc etc. What I'm used to though is unraid which gave a selection between the different options. Threadripper 1950X is actually extremely good with temps and power (when it came out anyway) so this is waaaay off. Also this is a fundamental part of the OS really so think I'll search / look for a ticket on it.

 

[Kris Moore]

I've just come here because I'm trying out TrueNAS SCALE and found it's misrepresenting the CPU temp - 70 degrees C at idle / 0% load. I'm already aware of the Tdie etc etc. What I'm used to though is unraid which gave a selection between the different options. Threadripper 1950X is actually extremely good with temps and power (when it came out anyway) so this is waaaay off. Also this is a fundamental part of the OS really so think I'll search / look for a ticket on it.

If you don't find one already existing please file a new bug ticket, that's something I'd think we'd like to fix before proper release since we don't want to be scaring folks with temps that are too high on the dashboard.

 

[marshalleq]

Fantastic, thanks will do. I'm flooded with little things as expected when trying out something new that isn't fully released yet. Will do shortly.