Nanotechnology is a rather loosely defined term, which many people have
come to use differently, depending upon to whom they are applying for
funding for their pet project. Drexler, who has inspired a large number of
researchers around the world, seems to stick to the molecular assembly
approach that Charles mentions below, whereas actual research varies according to
a very broad range of approaches, many of which have been a part of
science for decades and recently renamed nanotechnology. I agree with Charles'
comment in that I don't see how you manage the energy issues associated
with the Drexlerian approach, but that's someone else's problem. I don't see
how it would work fast enough for rapid prototyping/manufacturing to be
On the other hand, if inexpensive means for producing large volumes of
nanostructured starting materials having predictable properties could be
produced in volume and put into a machine that could manipulate them in a way
to cause them to self assemble into a structure with functional
properties, we might have something. I expect the first applications to be in niche
markets and to have very expensive assembly processes.
I'm not sure to which IBM development Elaine refers. If Elaine means the
experiments with carbon nanotubes as circuits, that is indeed exciting
stuff, which might be available to replace the current silicon technology in
15 or 20 years. The IC manufacturing industry will be reluctant to change
materials until they have to and a lot of tricks remain to be made with
silicon and the software to operate chips that will extend Moore's Law for
that length of time, if not in size, then in cost of operations. The
biggest blocks to silicon may not the technical challenges, but rather the
financial ones of paying to build fabs that cost 2 or 3 billion dollars each
when it's not clear who will want to buy those chips in sufficient volume as
to allow the chip makers to amortize those fabs. You might ask the same
question of nanotech: Who wants it in sufficient volume to pay for the cost
of producing the materials and the machines to produce products that way?
The use of CAD will be important. Such small devices will have to be
designed in CAD, in order to create the instructions for a machine or swarm of
nanomachines to build them. The CAD software will also need to
incorporate the kind of material property data that the IBMs of the world will need
to accumulate over the next 15 years, in order to design the devices they
want to build. The properties of carbon nanotubes, for instance, vary
considerably and many of the theoretical properties are unproven in reality.
Once the properties of a specific type of nanotube are known, someone needs
to invent a cost-effective process for producing commercial quantities of
that particular type of nanotube.
Many new discoveries have yet to be made by the scientists, many new
inventions are yet to be made by the engineers, and many new desires have yet
to be foreseen by entrepreneurs. All of that will likely take much longer
than most people think.
SRI Consulting Business Intelligence
+1 650 859-4350
Charles Overy wrote:
>You may have noticed that Scientific American this month has a number of
>articles on nanotech. There numerous links in the mag as well as at:
>Having worked on "Star Wars" misslie defense in the first Regan boom,
>extremely dubious of "national initiatives" particularly when they come
>pushing money at ideas that are very ill defined projects. The Nanotech
>initiative seems too close to me. However if this is the only way to
>basic research dollars out of a republican administration then so be it.
>Hmmm this probably was not the discussion you were hoping to inspire...
>Micro assemblers that can put things together molecule by molecule a la
>Drexler seem to be a pipe dream.
>However the technology at larger orders of magnitude seems to be very
>exciting. To me one of the very interesting technologies is self
>replication. At what point can a machine make a copy of itself so that
>can get exponential growth in output. Until that happens you have to
>very expensive and specialized machines that are dedicated to making
>specific lines or types of components. A machine that could operate at
>scale fine enough so that the generic properites of its source materials
>could produce a wide variety of functional outputs would be very
>For instance, what about an RP type machine that could opperate at the
>molecular level with carbon. It should be able to produce a part that
>carbon nanotubes, buckyballs, loose carbon powder, and possibly diamonds
>from one "feedstock" However, managing the energy input and molecular
>bonding is perhaps impossible if not very difficult.
>One thing is for sure. The parts won't use STL format, hooray!
>From: email@example.com [mailto:firstname.lastname@example.org]On
>Behalf Of jim quinn
>Sent: Monday, August 27, 2001 11:49 AM
>Subject: re: nano and RP
>re: nano/RP thread below
>If you are talking about nano-manufacturing, then there is
>a long way to go. Most of the current MEMS work is still meso
>or micro at best. Even for litho, 0.1microns is 100nanometers.
> > From email@example.com Mon Aug 27 13:27:24 2001
> > To: RPML <firstname.lastname@example.org>
> > From: Elaine Hunt <email@example.com>
> > Subject: a new thread
> > I am interested in hearing opinions about how nanotechnology and RP
> > one day combine or merge and impact product development. IBM's latest
> > development in transistors is a great example of size reduction and
> > development will lead to a new breed of computers. What do you see
> > happening in the area of CAD for nano objects, hardware
> > Elaine
For more information about the rp-ml, see http://rapid.lpt.fi/rp-ml/
This archive was generated by hypermail 2.1.2 : Fri Jan 04 2002 - 09:57:43 EET