LinkedIn emplea cookies para mejorar la funcionalidad y el rendimiento de nuestro sitio web, así como para ofrecer publicidad relevante. Si continúas navegando por ese sitio web, aceptas el uso de cookies. Consulta nuestras Condiciones de uso y nuestra Política de privacidad para más información.
LinkedIn emplea cookies para mejorar la funcionalidad y el rendimiento de nuestro sitio web, así como para ofrecer publicidad relevante. Si continúas navegando por ese sitio web, aceptas el uso de cookies. Consulta nuestra Política de privacidad y nuestras Condiciones de uso para más información.
“The continuing rise of Si based semiconductors is perhaps the major
technological fact of the past ﬁve or more decades. Silicon-based technology is
a “general purpose technology” [Bresnahan and Trajtenberg (1995)]
underlying much of the improvement in information storage, information
transmission and computation since the 1960s and some have argued
[Brynjolfsson and McAfee (2014)] that it is the most important general-
purpose technology ever. From 1968 to 2005, the number of transistors sold
for use has increased by 10^9; by 2005 there were more transistors used then
printed text characters (Moore, 2006)! However, the industry revenue per
transistor has fallen almost as dramatically (Moore, 2006) as has the amount
of material needed make a transistor. Nonetheless, the usage of silicon has
grown signiﬁcantly since 1970. We ﬁnd it has grown by 345% over this period
but also ﬁnd the growth is less than GDP growth (472% in the same period)
and that much of the growth of Si usage is associated with non-electronic
applications. This growth would be 10^5 (or more) times as high 2005
transistor used as much Si as one manufactured in 1968 showing the
importance of the profound change in “materials efﬁciency” this technological
Magee and Devezas (2017), A simple extension of dematerialization theory
Moore’s law postulates the doubling
of the number of transistors
crammed onto computer chips every
Densification and the increase of the
number of transistors account for
the increase of computation power.
On the left is represented the
number of computations per second
per computer which doubled every
year and a half between 1975 and
Credits: "Silicon Quantum Integrated Circuits.
Silicon-Germanium Heterostructure Devices:
Basics and Realisations”
E. Kasper and D. J. Paul Springer, Berlin, 2005
Most advanced transistors comprise
merely a dozen of atoms…
In June 2017, IBM announced about
a breathrough with 5nm processors
built around a new architecture
Beyond the 3 atom limit, one would
need to build single-atom
Fizaine, F. and Court, V. (2015) “Renewable electricity producing technologies
and metal depletion: A sensitivity analysis using the EROI”, Ecological Economics,
110, pp. 106–118. doi: 10.1016/j.ecolecon.2014.12.001.
• Solar panels are a renewable energy that requires oil, coal and
other non-renewable resources…;
• Relations between metals have not really been taken into
account as of now: either as regards the extraction side of
things (for an exception, see Fizaine: F. (2014) Analyses de la
disponibilité économique des métaux rares dans le cadre de
la transition énergétique. Université de Bourgogne, chapter 4,
https://tel.archives-ouvertes.fr/tel-01127141/,) or their use in
production (batteries need both lithium and cobalt);
• Stress on resources is an additional geopolitical issue (we were
used to waging wars for the control over oil, we may now
witness wars for the control of the elements of the periodic
The “inductive argument”: e.g.: "in the past, it was forecast that oil
would become less available and it’s still here so there is nothing new
under the sun."
• Yet: Optimization vs. The rebound effect and physical limits;
• Yet: Dematerialization as an expected result of technological
progress and optimization vs. studies that put forward an inductive
argument against the expected benefits of dematerialization
(Magee and Devezas 2017, A simple extension of dematerialization
theory: Incorporation of technical progress and the rebound effect:
“the major empirical ﬁnding reported here [is that] direct
dematerialization due to technological progress will not occur”);
The inductive argument thus proves to be a counter counter
Where is research heading? (before
• Spintronics? Photonics? Quantum computing?
• Biomimetics/Neuromorphics (Memristor,
Sensible Machines) ?
• New architectures (Adiabatic
Computation)/New logics/New languages?
• Towards the search for a different material
basis (carbon-based microprocessors)?
“I implore my audience: Study physics. Help invent new kinds of
nanodevices with high adiabatic energy coefficients. Design,
build, and empirically test high-quality ballistic oscillators,
interacting with quasi-static logical states, driving adiabatic
transitions between them. Systematically find and eliminate
sources of dissipation in your prototypes, one by one. Extend
your designs to larger and larger scales of complexity, with larger
and larger logic blocks ever more tightly and precisely
synchronized. Design fully-reversible architectures, languages,
“Approaching the Physical Limits of Computing”,
Michael P. Frank
• The dualistic view is still prevalent in CS (especially
where the discipline is mainly focused on
algorithms and formalisms). CS understands itself
has entertaining a relationship that is closer to
maths than physics.
• Turing : “From the point of view of the
mathematician being digital should be of greater
interest than that of being electronic” (Lecture to
the London Mathematical Society on 20 February
I stole this quote from José Halloy as well!
Where is research heading? (after the
The ontological issue with digitality:
the overcost of an “ideal” world
Crédits: Brian Cantwell Smith
“It is this ability to ceaselessly cleanup after its own noise that so
powerfully enables computers to seemingly sever their dependency on
physical processes that underlie processing, storage, and connectivity.
Yet the physical characteristics of a resource (be it computation,
storage, or networking) cannot simply be transcended, and noise can
only be conquered at the expense of other resources. For example,
manufacturers must design electronic circuits using a voltage
differential between 0 and 1 broad enough to ﬁght off interference by
galactic cosmic rays (“single event effects”), at the cost of increased
power consumption (May & Woods, 1979); error-correcting codes,
widely used to protect against transmission interference, result in both
data expansion (and thus, reduced capacity) and increased processing
load. In the latter case, designers will choose among different codes
according to both the expected proﬁle of the noise (frequency,
intensity), and the resource trade-offs. Once again, then, independence
from the material can only be obtained at the costs of certain trade-
Jean-François Blanchette (2011), “A material history of bits”
The digitality “sandwich”
Noise control (increased power
May in turn deteriorate the
How about the Web then?
• We can imagine a Web disconnected from the
Internet (y concluding a Semantic Web). A low
tech or « mixed tech » Web.
Cf. the « downscaling the Web » initiative.
• See also the mesh systems which allow to
locally connect computers.
Shift from the Web We
Want to the Web We
A Web of transition –
not necessarily an
• Of course, we should also avoid any “collapse-
porn effect” and treat catastrophes as mere
opportunities to innovate.
• Thus, collapse informatics and post-colonial
computing (see for instance Philip, Irani and
Dourish (2012) ‘Postcolonial Computing: A
Tactical Survey’, Science, Technology & Human
Values, 37(1)) need to quickly converge.
• We have still have the means to work with high tech,
and do research, how are we going to use that limited
amount of time?
• And, conversely, are we going to avoid doing things
despite having the means to do so?
Designer Tony Fry is asking us to avoid “defuturing”.
But to do so we need to actively defuture unsustainable
futures and projects.
Time is of the essence: what can we do? what should
we (not) do?
• Futures: multiple and asynchronous;
• “L’avenir” synchronises diverging futures;
• Our revolutions may be temporary;
• The end of linear and cumulative progress;
• Our (dominant) technology/infrastructure/research
is not durables but we do inherit it: how are we going
to mobilize it before the effects of the new era
already set in motion (the Anthropocene) are being
• We shall let go of past visions of the
future in the 21st century especially in
• We should be ready to live in the ruins
of the digital world that we produce on a
• We need to imagine a cyberpunk future,
comprising « high tech », « low tech »,
alter-tech (permaculture ?), ruins, etc.
That’s already the world we live in! “The
future is already here — it's
just not very evenly distributed”, once
famously wrote William Gibson. It’s not
just true of progress but also of collapse;
• We need to ask ourselves: « what
models should we
"It’s not climate change – it’s everything
"The starting point is demand reduction.
Turn it off."
Last quote I stole from José Halloy!