Re: nano-assembly and computing

From: Robert J. Bradbury (bradbury@aeiveos.com)
Date: Fri Mar 24 2000 - 11:36:12 MST


On Fri, 24 Mar 2000, matthew gream wrote:

> Parallel to the development of nano-tech assemblers are quantum computing
> systems. I am wondering whether it is expected, or feasible, for nano-tech
> assemblers to create quantum computers ?

It is unclear (to me) at this time whether quantum computers have any uses
other than for producing rapid solutions for a limited set of problems.
I know of nothing that has equated a quantum computer to a turing machine,
which if I recall can compute anything that is computable.
[Someone correct me if this is wrong.]

Nano-assembly can build anything that can be assembled (a different problem
from computed). The current problem with quantum computers is that you
need a bench full of equipment to get a couple of qubits. Nanoassembly
might shrink this somewhat but how much remains very unclear.

>
> For example, quantum computers could be created as stand alone entities
> in themselves, or they could be embedded within other media/materials.
>
In the models I've seen so far you are always going to need some large
amount of matter to create the necessary entangled quantum states that
allow quantum computers to work.

> In the latter case, consider the development of a reconfigurable life form,
> where the operating intelligence is distributed throughout the material.

Distributed intelligence exists already, earthworms, I believe are examples.
You could probably even say that bacteria operate this way. I would argue
that an amoeba is a fairly reconfigurable life form, though not in a very
general purpose way since it can't disassemble and reassemble itself to
be something different. It may be true that ultimate reconfigurability
is impossible (once you disassemble yourself into atoms its kind of
hard to put them back in any way other than the few limited paths
allowed by molecular forces).

> There is no central brain, as there is with existing life forms.

But you want concentrated intelligence with low interCPU delay times
and high interCPU bandwidth. Eric's 1 cm^3 diamondoid rod-logic computer
equal to 10^5-10^8 human brains is perfectly sufficient for this and operates
on well known computer principles.

When someone can show me a quantum computer with as many effective
bits as Eric's computer, that even using the measly power equivalent
of the human brain(10W) is pushing through about 10^21 bits/second,
then I'll consider them as potentially useful.

With QC, not only do you have the problem of how many bits it has,
you also have the problem of the machine cycle time. You may be able
to get the answer out in nanoseconds, but it may take you hours to
reset the machine so you can execute another "instruction".

> This is Terminator 2 imagery obviously, but the science fiction
> looks like it could be created by the reality of nano-tech assembly
> and quantum computing.

T-2 can be created with standard nanotech, no QC required.

Robert



This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 15:27:36 MST