Eugene Leitl (
Thu, 31 Jul 1997 23:42:55 +0200 (MET DST)

On Thu, 31 Jul 1997, Hagbard Celine wrote:

> [...]
> Can any of the techno-savvy list members explain how a powerful enough
> computer can be installed in such a (presumably flying) mini-robot? I

Apparently you missed the aerobot thread. There are now few-10 g
semiautonomous microremotes in development (quite recently there has been
a conference on them -- military recog missions (even within buildings)
and antipersonnel being obvious applications). According to web lore
(truly fascinating, btw) a number of hobbyists are building quite powerful
jet turbines from simple ICE turboloaders, a recent article has shown that
gas turbines scale down very well, predicting quite ridiculous weight/lift
ratios. Using lithography on SOS (silicon on sapphire) wafers, I guess an
array of microturbines cum according fuel lines/pumps and steering logic
can be built. Unidentified flying wafers, so to speak. Inertial guidance,
CCD and GPS are more than sufficient for navigation.

> suppose this question applies to all other forms of nano-robotics
> (assemblers and disassemblers, etc.) as well. Would nano-scale

Molecular system autotreplication by means of machine-phase chemistry
might well be impossible. Don't hold your breath yet.

> processors be small/powerful enough? What effect, if any, does the
> Beckenstein bound have on this?

None. Bekenstein bound predicts the upper achievable limits of computation
a given material system is capable of. Which are ridiculously high. Real
systems only survive tiny amounts/velocities of bit tweaking before
turning into plasma. A recent article I've read predicts several TBit
storage (currently 4 GBit, NEC). Professor Hartnagel's QM devices (100x100
quantum well array on few nm^2) could lead to 10^4 bit/1 um^2 storage
density in 10/15 years. This is the end of the line for flat circuits at
RT (but this still a lot higher than I expected). Above processes are
implementable in diverse III-V, and plain SiGe. Ebeam and AFM litho as
well as heavy ion track etching are being proposed.

icbmto:N 48 10'07'' E 011 33'53''