[p2p-research] from an energy standpoint, distributed networks are just as > wasteful as centralized cloud computing

[Andy Robinson]

A few points about this - made without much technical knowhow or statistical
details...

First point - computer running energy is peanuts in energy terms, compared
to things like heating, lighting, air conditioning, cars, planes, etc.  (For
comparison:  computer monitors actually use much more energy than hard
drives).  Ecologically, small-scale renewable energy use for low energy
consumption of this type should not be difficult; the problems are with the
big energy hogs.  For example, it is possible to run some computers (I think
the OLPC is a case in point) on solar energy, solar rechargeable batteries,
and even hand wind-up.  Since doing things on computers may save energy
which would otherwise have been used going out to obtain things or
transporting them around (e.g. car journeys to the cinema or video store, or
in extreme cases, regular transit to work), their carbon effects are by no
means necessarily negative.

This isn't to negate the problems in the intersection of computers with
ecological issues.  On a broader level, computers come into sustainability
problems in relation to components such as silicon (and coltan?) and
reliance on mass industrial production - which might well be casualties of a
transition to a peer-to-peer world.  Computers are also generating huge
amounts of "e-waste" due to rapid turnover.  I'd suggest both of these
problems could be solved by adopting something more like an open-source
ethos in general, avoiding built-in obsolescence and relying on modifiable
architectures (which takes hardware skills), so that existing computers
could be kept going indefinitely rather than being replaced every few years
(the latter an effect in part of market pressures - computers are designed
to be discarded rather than updated, this makes manufacturers more money).
Anyway...  I'd suggest this is where we should be directing our concerns
regarding ecological effects of computers.  It shouldn't be too hard to get
to the point where we have renewably powered, (relatively) easily updateable
computers, and the latter would certainly be on the side of open source (and
Macs).

Second point - the internet, like most functioning assemblages, does not use
"money" as such, but rather, specific inputs which have been commodified
today.  The crucial issue is whether these specific inputs are necessarily
scarce and whether they can be made sustainable.

Third point - this argument is discussing technical effects of different
models of server functioning given current levels of technology and
availability.  Price of computer technologies seems to be rather erratic and
not at all a matter of resource inputs - it's mostly about extraction of
rent on new (not yet diffused) or protected (copyrighted etc.) items of
hardware and those which have not yet (but may later) become widespread.
And this in a context where a large proportion of hardware development is
done for profit in a market dominated by subjective expectations and
convenience (cf. the persistence of Microsoft's market dominance as OS
despite clearly not being the "best").  Have we got any reason to assume
that the lower-energy-cost servers used by big centralised companies are
essentially connected to big centralised companies?  Aren't they likely to
be part of basic computer infrastructure or available cheap on the web in,
say, twenty years time, like scanners, modems and laser printers before
them?  Or, if they reduce energy use a lot, wouldn't they be mass-produced
if energy costs were pushed to the point where buying them became
money-saving for average users?  Or, isn't someone likely to invent a
portable, low-cost equivalent soon enough, a few years behind the original?
If not - is the fact that they haven't or won't a consequence of technical
difficulties or just of the dominance of big companies in hardware R&D?

bw
Andy
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listcultures.org/pipermail/p2presearch_listcultures.org/attachments/20090714/fedc51ba/attachment.html>