Seasons of Jason
What kind of power consumption tests would you like to see across #Windows and various #Linux distros?
Some ideas I have so far:
1: Proprietary vs #FOSS: Outlook vs Thunderbird, MS Word vs @libreoffice Writer, etc. (Have more pairs to suggest?)
2: Power savings when #gaming at native 4K versus using AMD FSR.
3: Idle/Load power usage of KDE, GNOME, Xfce, MATE, etc.
4: Power usage across various form factors (Tower/Desktop PC vs Laptop vs #SteamDeck)
YOUR IDEAS HERE 👇
Michael
@killyourfm @libreoffice Try to compare sleep state power consumption between different OS. Compare idle consumption.
Seasons of Jason
@mschfr Excellent idea!
Seasons of Jason
@toad Mmm, great suggestion. Consider that your CPU needs to work harder to decode content at higher resolutions, and I bet it does make a somewhat significant difference.
It's added to the list.
Toad :kaos:
@killyourfm I don't know if it's useful, but I would like to know if the power consumption while watching a 4k/HD/SD streaming movie. 4K is a great bandwidth consuming (not eco friendly), I'm curious if it has also an impact on POWER consumption
Seasons of Jason
@baldpolnareffart Automating is way beyond my skillset. Is that something you can do?
Bald Polnareff Art
@killyourfm @libreoffice I think we should collect all these ideas and then figure out a way to automate the testing under very specific conditions. What I mean is that we should probably try to change one variable at a time, we could probably come out with some sort of benchmark performing some sort of standardized test, ideally
Seasons of Jason
@baldpolnareffart @phoronix I'm not POSITIVE, but I think Phoronix Test Suite relies on the sys.power system sensor being available. I have two systems right in front of me where it's not.
You can find our your system's sensors by running:
$ phoronix-test-suite system-sensors
Bald Polnareff Art
@killyourfm Not really, at least in a reasonable amount of time. I wonder if someone like @phoronix could do something similar, it would be interesting. Then we could easily collect more data in a standardized way and do a better quality analysis
Seasons of Jason
@baldpolnareffart @phoronix I have two devices that would be ideal candidates: one is a Thelio, and the other is a 17" laptop with a dedicated Nvidia GPU. (The other laptops I have just don't pack a good gaming punch, and personally that's something I want to test extensively).
I'll look into their sensor capabilities this weekend.
PS: sys.power availability makes this possible -> https://www.forbes.com/sites/jasonevangelho/2019/05/06/benchmarked-the-most-power-efficient-ubuntu-19-04-flavor-will-surprise-you/?sh=2cbcf426e341
Bald Polnareff Art
@killyourfm @phoronix
Do you have a machine that you could exclusively dedicate to testing? I would focus on a single device first, isolating variables would be way harder if we started randomly collecting data from different machines.
Oh, another idea to add to the notes, running a test (any, once we define something), both on an SSD and a hard drive, those spinning disks should be more power hungry if I'm not mistaken
Jyn Erso Lives!
I was just coming back to this and wondered how the SteamDeck compared to the others? Can I game the whole day in good conscience? 😉
Seasons of Jason
@jynersolives I haven't had time to do much testing for various reasons. But I promise to get around to it.
GNU social JP is a social network, courtesy of GNU social JP管理人. It runs on GNU social, version 2.0.2-dev, available under the GNU Affero General Public License.
All GNU social JP content and data are available under the Creative Commons Attribution 3.0 license.