CPU temperature.

Sidney_Kotic

unread,

Jan 8, 2022, 5:07:22 PM1/8/22

to

Recently I put together a little computer and installed Leap 15.3 on it, it's
patched completely.
uname -a
Linux opihi 5.3.18-59.37-default #1 SMP Mon Nov 22 12:29:04 UTC 2021 (d10168e)
x86_64 x86_64 x86_64 GNU/Linux

The motherboard is a ASRock B550M-ITX/ac and the CPU is a AMD 5700G.

When I enter the command sensors it replies
nvme-pci-0500
Adapter: PCI adapter
Composite: +35.9°C (low = -20.1°C, high = +74.8°C)
(crit = +79.8°C)

iwlwifi_1-virtual-0
Adapter: Virtual device
temp1: +53.0°C

Which isn't particularly informative. Since the numbers are pretty static.

Whereas another computer, a old Dell with a Intel G3220 CPU, says
coretemp-isa-0000
Adapter: ISA adapter
Package id 0: +56.0°C (high = +80.0°C, crit = +100.0°C)
Core 0: +56.0°C (high = +80.0°C, crit = +100.0°C)
Core 1: +53.0°C (high = +80.0°C, crit = +100.0°C)

acpitz-acpi-0
Adapter: ACPI interface
temp1: +27.8°C (crit = +105.0°C)
temp2: +29.8°C (crit = +105.0°C)

dell_smm-virtual-0
Adapter: Virtual device
Processor Fan: 1320 RPM
Other: +42.0°C

Which is certainly more specific, particularly the Core temperature. Which is
what I'm concerned with since I push the CPU's hard.

Is there a patch, package, or command sequence, available which will tell me the
specifics about the AMD core temperatures?

Malcolm

unread,

Jan 8, 2022, 5:19:20 PM1/8/22

to

Hi
Have you run sensors-detect --auto to configure?

--
Cheers Malcolm °¿° SUSE Knowledge Partner (Linux Counter #276890)
Tumbleweed 20220106 | GNOME Shell 41.2 | 5.15.12-1-default
HP Z440 | Xeon E5-2690 V3 X24 @ 2.60GHz | AMD RX550/Nvidia GT1030
up 1 day 5:37, 2 users, load average: 0.26, 4.92, 5.49

Sidney_Kotic

unread,

Jan 8, 2022, 5:49:39 PM1/8/22

to

On 1/8/22 15:19, Malcolm wrote:
> Have you run sensors-detect --auto to configure?
>

I have now, and have lots of new stuff to look at.

I'll up the workload and watch it, specifically the one that says
CPUTIN: +67.0°C (high = +80.0°C, hyst = +75.0°C) sensor = thermistor

At the moment I'm only running it at 50%, I'll kick it to 75%.

Malcolm

unread,

Jan 8, 2022, 6:06:31 PM1/8/22

to

Hi
You can configure the sensors to be more meaningful to you, rather than
CPUTIN if you like?

--
Cheers Malcolm °¿° SUSE Knowledge Partner (Linux Counter #276890)
Tumbleweed 20220106 | GNOME Shell 41.2 | 5.15.12-1-default
HP Z440 | Xeon E5-2690 V3 X24 @ 2.60GHz | AMD RX550/Nvidia GT1030

up 1 day 6:24, 2 users, load average: 0.04, 0.12, 0.52

Sidney_Kotic

unread,

Jan 8, 2022, 6:41:59 PM1/8/22

to

On 1/8/22 16:06, Malcolm wrote:

> You can configure the sensors to be more meaningful to you, rather than
> CPUTIN if you like?

At the moment that's acceptable. Once I decide to follow it I'll plug this into
my script:
Whew=`sensors | grep CPUTIN | cut -d "+" -f 2 | cut -d "." -f 1`
Later I check the value in $Whew.

Kicked the CPU utilization from 50% to 75% and not much is changing. CPUIN has
gone up 2C as has "Composite".

Carlos E.R.

unread,

Jan 9, 2022, 10:39:13 PM1/9/22

to

You can use "gkrellm" to see all those values.

--
Cheers, Carlos.

Andrew

unread,

Jan 10, 2022, 5:22:22 AM1/10/22

to

Agree on that. With gkrellm you can also see disc traffic (composite or
split by disc) and network traffic (ditto). You can also get a figure
for the cpu fan but that does not work on my hardware. Open the
"configuration" and experiment, changes are reflected in the display
immediately.
The displays for Mem usage and for Swap are not very visible, but I open
a terminal window and type "free" if that looks to be a problem.
I also have gkrellm running all the time with the "always on top" set.