Single-core Intel Atom Netbook (investigation) 2: What's the effect of removing 'Conky' on CPU & RAM activity in Manjaro-i3 ?

Thought I'd continue to share some further data (might be useful to someone), collected while I was investigating whether CPU & RAM activity in the i3 window manager would decline if conky was removed.


These data show that removing conky in Manjaro-i3 (version 18.1.2) resulted in idle CPU activity being reduced by 59%, whereas idle RAM usage fell by around 4%.


I recommend to try the commmand sudo ps_mem. This way you won't need to guess anything about RAM (apart from which share of the shared RAM each program uses).
But I like your visualisations anyway!

Conky isn't as benign and 'lightweight' as you're led to believe.

It uses ~ 2% of all 12 cores when it's running on my machine.

In OP's case, I flat-out wouldn't run it.

1 Like

Na conky is dependent on the user config, you want bells and whistles, or fancy lua effects it will use a lot, if you know how to code conky it uses next to nothing on my machine 2mb ram and virtualy 0 cpu.
Don't blame conky you the user are in charge of it,

1 Like

Many thanks for the very useful tip re. use of ps_mem :sunglasses:
The output in terminal is listed as:
Private + Shared = total RAM used.

I assume 'shared' is the RAM used for shared dependencies with other programs or apps, and 'private' is the RAM only used by conky: is that correct ?


It also depends on what you monitor and with which frequency. There was a time I used a fancy conky script on my KDE. Well, that didn't last long because of the needed CPU. There was just to much to monitor and the required frequency was to high.

Interesting observation. Given the number of cores you have, I'd imagine though that the extra 2% CPU activity is not going to make any difference to your machine's overall performance.

Just looking at the single-core Intel Atom system I have (#post 1), conky appears to cost around 2.3% in CPU activity (close to what you observed for a 12-core system!), and around 9 MB of RAM (broadly agreeing with the value given by sudo ps_mem - see eugen-b in #post 2).

However, when I check how efficient the system is, in installing the linux419 package & subsequently updating the multiboot grub-bootmenu, I do see a slight increase in performance (i.e. processing time is 1.3% less), but this doesn't appear to be significant:

With i3wm conky ......... 364 ± 4 seconds
Without i3wm conky ....359 ± 3 seconds
(n=3 in both cases)

Yep I think we have all been their try to have the greatest conky of all time i know i did. even to time and date why have that when its already in the panel,
conky can be lean as can be its all down to the user on a 4 core i5 i never see more than 1% on 2 of the cores with a simple conky no conky manager needed.


Forum kindly sponsored by