Permacomputing

659 readers

2 users here now

Computing to support life on Earth

Computing in the age of climate crisis is often wasteful and adds nothing useful to our real life communities. Here we try to find out how to change that.

Definition and purpose of permacomputing: http://viznut.fi/files/texts-en/permacomputing.html

XMPP chat: https://movim.slrpnk.net/chat/lowtech%40chat.disroot.org/room

Sister community over at lemmy.sdf.org: !permacomputing@lemmy.sdf.org

There's also a wiki: https://permacomputing.net/

Website: http://wiki.xxiivv.com/site/permacomputing.html

founded 1 year ago

MODERATORS

schmorpel

Comparing the environmental costs of buying a brand new high-efficiency Mini-lab computer, instead of a used less efficient one. (www.youtube.com)

submitted 8 months ago by ProdigalFrog to c/permacomputing

6 comments fedilink hide all child comments

I thought this video was rather interesting, because at 12:27, the presenter crunches the numbers to find out how many years it would take for a new computer purchase to be more environmentally friendly (in regards to total CO2 expended) compared to using a less efficient used model.

Depending on the specific use case, it could take as little as 3 years to breakeven in terms of CO2 if both systems were at max power draw forever, and as long as 30 if the systems are mostly at idle.

you are viewing a single comment's thread
view the rest of the comments

[–] tinspin 2 points 8 months ago* (last edited 8 months ago)

I think the most important element is longevity, and for that you need passive cooling that stays under 50-60 degrees Celsius at full blast. You also need disk solutions that are easy/cheap to replace and sturdy. That means old SD cards (SLC or MLC in that order) or old SSDs at ~65nm with 100.000 writes per bit. Modern NAND only has 10.000 WPB.

I still have my 45 nm Atom D510MO running 24/7 since 2010 in a passive heatpipe+sink case and 14 nm NUC + Atom 8-core in Streacom passive cases since 2014 and 2021 respectively. I'm waiting for the SSD in the NUC to fail. The recent Atom runs on X-25E that I know will outlast me.

N100 is 3x the performance per watt of the Raspberry 4, but it also costs 3x. I prefer the lower cost of the Raspberry 4 that allows more redundance even if that means more "garbage" has to be manufactured until that redundance is applied. Combined with cheap heatsinks and SD cards I already have.

The low power server of excellence is the Raspberry 2 at 2W TDP; I can have that running on solar only with a 100Ah battery bank... I think it would stay on a week without sun. The only 32-bit ARM multicore processor with vanilla linux that will ever exist for eternity, with the reduced complexity that brings!