Ask HN: Do Windows clients cause more CO2, than Linux distributions?

https://news.ycombinator.com/item?id=26618796

by keks24 • 5 years ago

1 3 5 years ago

news.ycombinator.com

Hello,

this might be more of a philosophical question.

I mean an idling Windows (maybe) uses 4 GB of RAM just like that, whereas an idling Linux distribution only uses 300 MB; talking about my Gentoo Linux using AwesomeWM.

The more RAM or resources are used, the more power is used, right?

-Keks

Someone2• 5 years ago

All other things being equal, _if_ unused RAM used less power: yes.

Unused RAM doesn’t use less power, though. Writing and reading it do, but afaik, all current systems happily refresh the memory content of all RAM, including unused RAM, and that’s what eats the power in an idling system.

Installing less RAM will help, though. I think that’s one reason why Apple ships iPhones with less RAM than flagship Android devices.

keks24• 5 years ago

Unused RAM doesn’t use less power, though.

Interesting, I thought it would. Just like a written hard drives weights more, than an empty one:

https://www.ellipsix.net/blog/2009/04/how-much-does-data-wei...

Someone2• 5 years ago

I don’t think that calculation is entirely correct. Hard disks use error-correction schemes and codes to prevent long sequences of identical bits, so writing logical zero bits doesn’t imply writing “magnetic zeroes” (https://en.wikipedia.org/wiki/Group_coded_recording, https://en.wikipedia.org/wiki/Modified_frequency_modulation).

DRAM = tiny capacitors, so keeping more of them discharged would mean needing less leakage and thus less energy to top them up, though, so some gains could be made there.

Ask HN: Do Windows clients cause more CO2, than Linux distributions?

Related Stories