Hi everyone, is it worth using power-profiles-daemon in a AMD system?
What does it do under the hood?
I’m using a Lenovo Legion 5. Thanks in advance.
Hi everyone, is it worth using power-profiles-daemon in a AMD system?
What does it do under the hood?
I’m using a Lenovo Legion 5. Thanks in advance.
I’ve read that and the gitlab page of the project, however none says how does it work under an AMD CPU, and I find benchmarks very hard to find on that matter.
Thanks.
power-profiles-daemon
operates on p-states, which AMD CPUs have. So, yes
However, I found that I’m using the acpi-cpufreq
scaling driver by default, and not the amd_pstate
one. Will it still make a difference?
Thanks.
My bad, I did not do much reading before delivering that answer. Having now done over 20 minutes’ research on this, I legitimately don’t know. There is not much evidence out there suggesting either way, which makes me think that it does work fine on AMD because someone would have raised a fuss about it.
The Arch Wiki states that CPU governors are driver-agnostic, but I am unable to find whether or not power-profiles-daemon
directly uses governors. I assume it does, but I’m not good enough at reading C and don’t have a good enough understanding of governors to figure it out.
This comment implies that power-profiles-daemon
supports modern AMD CPUs on newer (6.0+ should be good) kernels, but the link they provide is dead so it is inconclusive.
Really, I would say the easiest way to get an answer is to run your own benchmarks. If it is working, you should notice a significant difference on CPU benchmarks between power saving and performance modes. On my i7-12700H, I can consistently get more than a 2x performance difference using sysbench --test=cpu run
.
Thanks a lot for you response, I will investigate further.
This topic was automatically closed 2 days after the last reply. New replies are no longer allowed.