[extropy-chat] Dark matter and ET

[Eugen Leitl]

On Tue, Jul 19, 2005 at 01:38:19AM -0700, Damien Sullivan wrote:

> Maybe, but saying that DNA itself is computronium is I think a bit much.  It's

Computronium is a molecular crystal made from computing cells, optimized 
for doing computation. DNA is nothing like computronium. Dry cold DNA is a pretty
dense storage medium, though.

> a nicely dense storage medium, and yeah a few problems have been tackled in
> the lab with naked DNA, but we're a ways yet from a general system.

DNA computes by matching complementary bases. It's a solvated linear biopolymer, 
requires lots of solvent to be able to diffuse freely. It's limited by viscous
drag, and by I/O issues, required to encode your problem, and decode it.

Electronic and spintronic systems are much quicker (ballistic electron and photon
transport), and can be much denser, since requiring no solvent nor bulky encoding/decoding
hardware.  

DNA might be useful to build computronium by way of self-assembly, but it's really
lousy for computation.
 
> > reason for gratitude. Inefficient? Show me an artificial solar power device
> > more efficient than chlorophyll and photosystems I & II. 
> 
> Well, depends on the purposes.  For generating electricity from sunlight you
> want photovoltaics, at 15-30% efficiency, or a solar-driven turbine at perhaps

There might be alternative ways of converting solar radiation, with efficiencies not
limited to ~30% by the band gap (though stacking tuned-gap cells could range higher)

	http://www.nrel.gov/ncpv_prm/pdfs/papers/54.pdf

> Now, what *is* biological computronium is neural matter.  I used to take for
> granted that it was crappy, what with using neurons instead of wires, even if
> they were self-repairing, and having unbelievably slow signals.

No, in terms of energy efficiency and functionality concentration a human or
an insect brain is very far removed from what we can currently do. 
 
> But I figure a brain has 1e6 to 1e8 times the raw computing power of a desktop
> CPU (or more accurately, would take that many CPUs to be emulated), while

We don't know what the equivalent is. There are no comparable benchmarks.

> using about as much power.  I'd thought CPUs massed a few grams, so the brain

CPUs alone are useless. You need to look at the total mass of a Blue Rack, included
the machine room and the air conditioning versus a ~75 kg human. Here the efficiency
becomes particularly abysmal.

> would be only 1000x as mass efficient, but on checking with friends I'm being
> given the idea that modern CPUs are in the hundreds of grams.  And brains are
> more resistant to EMPs, and have their memory built in (which is probably part
> of the efficiency).

Um, there are many reasons why our current systems suck. However, there are also
many reasons why a nanocomputer would beat, say, the bee brain by many orders
of magnitude, both in speed and efficiency.
 
> I'd wondered why a brain should be more power efficient, given that it's using
> whole ions as charge carriers, instead of light electrons.  And why neurons

This factor is negligible.

Sorry for the handwaving, writing a quantitative paper comparing CNS and
current hardware would be fun, but not something I'm paid for. If other parties
are interested, we might cook up a list of differences and similiarites, and put
some numbers to it, so the archives should get it. (Not that anybody can be arsed 
to read the archives, of course).

> didn't lay down conducting polymer to use fast electrons instead.  But then I
> remembered KE = mv^2/2.  Ions are 10,000x more massive than electrons... and
> being moved about 1e6 times more slowly.  I don't know how many ions are used,
> vs. how many electrons, but it's at least a sign that moving small things very
> quickly is in fact not the route to energy efficiency.
> 
> Pit about the lack of backups.

They weren't in the original specs.

-- 
Eugen* Leitl <a href="http://leitl.org">leitl</a>
______________________________________________________________
ICBM: 48.07100, 11.36820            http://www.leitl.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20050719/9ac93380/attachment.bin>