Re: Proposal for a new W3C CG: "Web We Can Afford"

Hi everyone, I think this is probably the original source, a conference paper from Gerhard Fettweis and Ernesto Zimmermann: https://mns.ifn.et.tu-dresden.de/Lists/CustomListDefinitions-RoadmapListInstance/Attachments/2/Fettweis_G_WPMC_08.pdf


It’s a short read, and the relevant passage is:

Currently, server farms and telecommunications infrastructure are responsible for roughly 3% of the world wide electricity consumption. If the present growth trend of 16% per year continues, as the increase of internet traffic[2]  and the number of mobile phone subscribers suggests3 , this consumption rises by a factor of 30 in only 23 years: the current level of world electricity consumption.

The overall growth trend seems to be extrapolated from two sets of studies.

1) Projecting a 40% annual growth in IP traffic:

[13] Atsushi Ogasawara “Energy Issues Confronting the ICT Sector”, Science & Technology Trends Quarterly Review, No. 21, October 2006
and
[19] Cisco Visual Networking Index – Forecast and Methodology, 2007–2012, available online from http://www.cisco.com/


2) extrapolating growth in mobile phones from growth in India
[14] Telecom Regulatory Authority of India, Press release available at http://www.trai.gov.in/trai/upload/PressReleases/511/pr22nov07no96.pdf


My impression is that this projects a worst case scenario that could occur if growth is consistent for the next 30 years, and the energy efficiency of web technologies stagnates over the same period.  The reality will probably be less extreme (one hopes!).

At any rate, I'm very pleased to see the creation of this group. One of my interests is material impacts of the web – a set of technologies that is so often presented as immaterial (e.g. in popular conceptions of "the cloud”) – and ecological impacts are a big part of this.

Regards,

Jack Jamieson
PhD Student, Faculty of Information, University of Toronto
jack.jamieson@mail.utoronto.ca<mailto:jack.jamieson@mail.utoronto.ca>

On Jan 25, 2016, at 12:37 PM, alexandre.monnin@web-and-philosophy.org<mailto:alexandre.monnin@web-and-philosophy.org> wrote:

Hi,

Le Mon, 25 Jan 2016 11:37:57 +0100, Brandt Dainow <bd@thinkmetrics.com<mailto:bd@thinkmetrics.com>> a écrit:

Hi - I am not sure the figures for energy consumption will stand close examination.  I have read the paper for which Alexandre kindly provide the link.  Its source for these numbers is an article in ZDNET, which is a commercial, not peer-reviewed, publication.  In  which they state:

"Dans 25 ans, si les technologies informatiques ne progressent pas, l’internet consommera autant d’énergie que l’humanité en 2008. C’est le calcul qu’a réalisé Gerhard Fettweis de l’Université de Dresde."

Funnily, I've discussed with someone from Ademe about this point and he confirmed it was a study from Dresden. I could ask him the exact source.

This said, I think part of the group interest could be to find means to improve the way we access the source behind such studies (that's just one example).

They also cite a number of studies making claims such as "one search of Google uses the same energy as one hour's use of a lightbulb."

Which I haven't mentioned myself ;)) Mainly because the group, again, should not be just about IT & energy to my mind.

The studies they quote are not researched, but uncited claims made by eco-groups, such as Green IT.  On a purely factual basis, such a claim is meaningless - what sort of lightbulb, halogen, neon, etc?  Of what strength?  ZDNet do not offer a formal citation for the Fettweis number, even though they provide formal citations for other documents.  I have looked at Gerhard Fettweis's page at the University of Dresden, where he is Vodaphone Chair of Mobile Communications.  There is nothing in his list of publications related to this issue.  His publications are all concerned with the technology of mobile phones.  Some of these do discuss energy efficiency, so it is possible he mentions this figure in one of those publications.  However, I cannot find any publication by him which looks like it could be a direct study from which this number is a research finding.  IF he is the source of this figure, then it looks more probable he is quoting another study.

On the basis of my research, I have to conclude that we do not have a proper source for this figure because it cannot be traced to any originating publication.  I therefore suggest it is an unsafe figure to use in any debate because it is easily challenged, but lacks any supporting evidence which could be used to defend it.

I would love to find proper evidence to support this claim so I could use it in my own research into the ethical implications of emerging ICT's, so if anyone can find a reliable source, I would be interested to see it.

Helping to find and reference such evidence could be part of our discussion...

Best,
A.

Regards,
Brandt Dainow
brandt.dainow@gmail.com<mailto:brandt.dainow@gmail.com>

https://www.researchgate.net/profile/Brandt_Dainow

http://www.imediaconnection.com/profiles/brandt.dainow



-----Original Message-----
From: alexandre.monnin@web-and-philosophy.org [mailto:alexandre.monnin@web-and-philosophy.org]
Sent: 23 January 2016 15:49
To: public-philoweb@w3.org
Subject: Re: Proposal for a new W3C CG: "Web We Can Afford"

Hi Henry,

Le Thu, 21 Jan 2016 20:58:13 +0100, Henry Story <henry.story@bblfish.net> a  crit:


On 21 Jan 2016, at 17:10, alexandre.monnin@web-and-philosophy.org wrote:


Here is the description of the group:

Most scientists now seem to agree that we've entered a new epoch dubbed
the "Anthropocene", where the environmental consequences of human
development have a tremendous impact on Earth's equilibrium. Those
effects are already set in motion and will have far-reaching
consequences in the coming years despite all the measures we could take
to mitigate them (considering we simply do not fail to take action).
While trying to avoid some of the consequences of the Anthropocene is
an issue that is well-worth striving for, another task would be to
reconsider the design of things at the time of the Anthropocene and
that includes the Web. For instance, a 2008 study by the University of
Dresden stated that if no measure was taken, the energy needed to power
the infrastructure of the Web in 2030 would be tantamount to the energy
consumed by humanity in 2008.

Is there a link to the study?

I quote a study from Ademe (the French angency for the environment) which
mentions it:
http://www.presse.ademe.fr/wp-content/uploads/2014/07/Web-Energy-Archive-Note-de-synth%C3%A8se.fr_2.doc


The agendas of the stakeholders who are trying to set the Web forward
in motion are mainly focused on adding new technological layers to the
existing ones. Yet, the logic behind these developments remains that of
tapping into unlimited resources, not limited ones.

It is true that the PCs behind Google running Linux were built for the
PC industry and have been well known to be the least energy efficient.
Still in this space there is huge opportunity for improvement.

True. The question regarding improvement is whether you improve to do less
or more (so that it may eventually lead to negative results).

When I was working at Sun Microsystems, we developed the T series of
microprocessors that were very efficient in energy
consumption. Sun argued that the energy saving alone could pay back for
these very expensive computers within one year
https://en.wikipedia.org/wiki/SPARC_T-Series

<https://en.wikipedia.org/wiki/SPARC_T-Series>

The latest of these now have 32 cores, 256 threads
https://www.oracle.com/servers/sparc/silicon-expression-of-oracle-strategy.html

<https://www.oracle.com/servers/sparc/silicon-expression-of-oracle-strategy.html>

The large cloud providers consume a lot of electricity, but they also
have staff to measure that electricity consumed, and the
money to optimize as much as possible out of it. Those optimizations
then find their way into software architectures that
can be more widely used in more distributed networks which can use solar
engergy, heat, etc to power the IoT of a home.

Henry, please note that this group is not about energy optimization in IT.
It's about assessing the impact of IT but more than that, about the kind
of IT we need (if any) in a world that will be forever changed in just a
few years, facing effect that are already set in motion.

I was happily surprised yesterday to discuss with lots of Inria
researchers who are more interested in this issue, which has to do with
notion of imminent collapse, that with the so-called ecological
"transition".

Lots of endeavors are currently focused on reshaping the Web into a
"Web we want", a redecentralized open Web fit for an enlightened
digital age. Those who advocate such an agenda and those who oppose it
generally both share a common assumption: that enlightened or not, the
future will be even more digital than the present. Yet, life at the
time of the Anthropocene, at least in the coming decades, might not
remain as pervasively digital as it is today. Other efforts that see
the ongoing battle for the decentralization of the Web as an
opportunity to  downscale  it (in particular in Africa) seem to be
aware of that. Maybe it's time to take into account other perspectives
on the future and concretely act towards building a sustain-able (Tony
Fry) Web. In other words, a Web We Can Afford. This group would like to
reconcile the development of the Web and an awareness to the
environmental issues by appealing to Web architects and designers,
eco-designers, activists, philosophers, social scientists, etc., so as
to make the issue a public one to begin with, before devising a set of
guidelines as a first step towards concrete action.

I think it's interesting. This may actually be a topic for the Web
Science group ( see Southampton ) to investigate.

Let's ask Les then! ;)

I have a feeling that this is a very holistic type of problem. Because
one cannot just look at the electricty consumed one has to look at the
efficiencies gained through that consumption. One has to consider
exponential improvements in efficiency at all levels, from moving stuff,
to capturing and transforming energy, to knwledge sharing, which itself
can advance even further the improvements in the other areas.

Odds are, for such improvements to have a positive effect, you'll need to
tackle difficult political issues. Optimization (which is not a silver
bulet) is not all there is to these complex questions.

Best,
A.




--


* Membre du collège d'experts Open Data de la mission Etalab du Premier Ministre
* Chercheur associé chez Inria (EPI Wimmics, Sophia Antipolis)
* Co-initiateur du projet DBpedia Francophone et SemanticPedia
* Docteur en philosophie à Paris 1 Panthéon -Sorbonne (PHICO, EXeCO) - Thèse sur la philosophie du Web : disponible et annotable sur http://philoweb.org<http://philoweb.org/>
* Co-chair du Community Group "Philosophy of the Web" au W3C
* Organisateur des "Rencontres du Web de données"

http://web-and-philosophy.org/,

Twitter : @aamonnz & @PhiloWeb,
PhiloWeb on Dailymotion, PhiloWeb discussion list @INRIA