The footprint of those is surely going to become an issue going forward, that's explains marquee pre-emptive actions to link them to renewables.
The question is whether intrinsically (and ceteris paribus) #decentralization consumes less (because traffic remains more localized)
@openrisk I'm also very interested in this question but I fear the comparison is all but trivial...
You would have to define very precisely your "ceteris paribus" in terms of features (e.g. downtime, backups...) and define how the "decentralized" platforms would be hosted (today, most Fediverse instances are not hosted on a RaspberryPi at home but by companies like OVH).
In the meantime, you may be interested in the work done by @greenfediverse if you dont' know it yet
@tfardet yep it seems highly non-trivial (thats why the qualification) but nothing important is easy 🙂
there are a number of factors in favor of digital decentralization (not an expert by any means), latency, resilience etc and lower ultimate resource consumption might be another one. There might be also some cons, eg less efficient use of hardware
thanks for tip will try to connect to @greenfediverse
There's also the fact that centralised platforms like FB and Twitter (and Gmail and...) process and analyse all the data flowing through them, for the purpose of creating profiles and targeting ads.
All other things being equal, doing less work saves energy.
@owl in that vain there is also the hierarchy of text -> audio -> video usage, where the later is promoted as it stimulates more strongly even though its information content is actually the lowest :-)
but this takes us a bit into the subjective domain of whether the nature of the traffic volume justifies the resource consumption
@openrisk @humanetech I’m no expert but this idea is interesting to me. I would hypothesize the opposite - I think it’s generally more efficient to share resources on one larger machine than divide it up - you would have more resources wasted by idling. This seems apparent with memory, but maybe there’s more nuance with electricity.
@jaredmoody @email@example.com its an optimisation problem of sorts and to solve properly one must indeed figure all the important elements. Losses in idling infrastructure would be definitely adding on the penalty side of extreme decentralization. On the other hand the gains would probably be quick to recover if one had, say, neighborhood shared clusters pooling compute for ~100 or so.
Another interesting angle is whether cooling is more efficient with less aggregation
@openrisk In my mind I associate data centres with being pretty efficient, even if that only happens at scale, as there are servers which do a lot of things in one centre. Someone mentioned that fedi servers do less work, but I wonder if fault tolerance changes anything. Centralised systems still have redundant hardware, and federated systems tend to replicate remote information on each server.
@iron_bug localized in the sense that most people would be exchanging information with others in their vicinity (say city level) and signing up to geographically close instances would have various advantages.
Storage and compute efficiencies are an important factor and maybe they trump any network efficiencies. But it may also be the case that an optimum of sorts is far more decentralized than current, especially if one takes into account other factors such as resiliency.
@iron_bug there is also the growing disquiet in some places about consuming energy (renewable / clean or not) in one spot to serve communication needs in remote places
Welcome to mastodon.green This server is for people in Europe, but you can connect with friends on any Mastodon server in the world.