Survey: provide your available CPU governors

[fulalas]

I would like you guys to run the command below and tell which CPU you have:

Code: Select all

cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_available_governors

This is part of a research I'm conducting to better understand the differences between kernel CPU governors and alternative solutions like power-profiles-daemon and TLP.

Thanks in advance!

[Rapha_]

Code: Select all

guest@porteus:~$ uname -mpi   
x86_64 AMD Athlon(tm) Dual Core Processor 4850e AuthenticAMD

2x 2500,00 MHz

Code: Select all

guest@porteus:$ cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_available_governors
ondemand performance schedutil 
ondemand performance schedutil

[Ed_P]

Code: Select all

guest@porteus:~$ uname -mpi
x86_64 11th Gen Intel(R) Core(TM) i5-1135G7 @ 2.40GHz GenuineIntel

guest@porteus:~$ cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_available_governors
performance powersave
performance powersave
performance powersave
performance powersave
performance powersave
performance powersave
performance powersave
performance powersave
guest@porteus:~$ 

[donald]

Code: Select all

i686 AMD Athlon(tm) 64 X2 Dual Core Processor 5000+ AuthenticAMD
guest@porteus:~$ cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_available_governors
ondemand performance schedutil 
ondemand performance schedutil 

[fulalas]

Thanks a lot!

[ncmprhnsbl]

Code: Select all

AMD Ryzen 5 2600
cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_available_governors
conservative ondemand userspace powersave performance schedutil 
(x12)

[M. Eerie]

AMD Ryzen 7

Code: Select all

conservative ondemand userspace powersave performance schedutil 

x16

[rych]

Code: Select all

root@porteus:~# cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_available_governors
performance powersave
performance powersave
performance powersave
performance powersave
performance powersave
performance powersave
root@porteus:~# uname -mpi
x86_64 Intel(R) Core(TM) i5-9500 CPU @ 3.00GHz GenuineIntel
 

[Vic]

Code: Select all

root@VICS-HP-PORTEUX-1:~# uname -mpi
x86_64 11th Gen Intel(R) Core(TM) i3-1115G4 @ 3.00GHz GenuineIntel
root@VICS-HP-PORTEUX-1:~# cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_available_governors
performance powersave
performance powersave
performance powersave
performance powersave
root@VICS-HP-PORTEUX-1:~#

[Rava]

Code: Select all

guest@rava:~$ uname -mpi
x86_64 Intel(R) Core(TM) i7 CPU       Q 720  @ 1.60GHz GenuineIntel
guest@rava:~$ cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_available_governors
ondemand performance schedutil 
ondemand performance schedutil 
ondemand performance schedutil 
ondemand performance schedutil 
ondemand performance schedutil 
ondemand performance schedutil 
ondemand performance schedutil 
ondemand performance schedutil 
guest@rava:~$ 

It is a 4 Core CPU and each core is able to run 2 tasks simultaneously, therefore the max CPU% would be 800%, but when asking for the Core's temperature only 4 are listed via sensors|grep Core :

Code: Select all

guest@rava:~$ sensors|grep Core
Core 0:       +69.0°C  (high = +84.0°C, crit = +100.0°C)
Core 1:       +69.0°C  (high = +84.0°C, crit = +100.0°C)
Core 2:       +70.0°C  (high = +84.0°C, crit = +100.0°C)
Core 3:       +74.0°C  (high = +84.0°C, crit = +100.0°C)
guest@rava:~$ 

The "high = +84.0°C" is always the same, that temperature is happening at initial bootup; after that temperature got reached the system starts the fan.

[nanZor]

Cool - I like LXDE and add the native cpu and governor app "cpufreq frontend" in the taskbar, where I can hover with the mouse. (Add-Remove Panel Items > ADD > CPUfreq Frontend) But here we go for this machine. I'll add other machines later..

Code: Select all

guest@porteus:~$ uname -mpi
x86_64 Intel(R) Celeron(R) CPU  N2940  @ 1.83GHz GenuineIntel

Code: Select all

ondemand performance schedutil 
ondemand performance schedutil 
ondemand performance schedutil 
ondemand performance schedutil

With the lxde panel app, I can hover and see the cpu frequency changing in real-time at idle or under a workload easily...

[nanZor]

Just an update - I found this handy cli one-liner to watch the cpu speeds. Yeah, Htop has that option etc, but this is always available:

Code: Select all

watch -n1 "grep \"^[c]pu MHz\" /proc/cpuinfo

There is a space between the p of grep and the backslash. And a space before /proc. Formatting sometimes makes that hard to see. Ctrl-c to quit obviously.

Interesting to see how you are doing at idle and under load with the various freq governers etc. -n1 is one-second. You can make it smaller, but then the command itself starts to make it a load!

I mostly use it to see if my cpu's actually settle down at idle, rather than staying hammered at full-speed.