On Mon, 10 Dec 2001, Dossy wrote:
> What James and I were discussing was at the microscopic level. I was
> suggesting that "if photosynthesis isn't efficient enough to power a
> small device, then perhaps that small device needs to be engineered to
> require less power while accomplishing the same task" ... and James is
> asserting that modern technology is already reaching (or has reached)
> their limits.
We need to be very careful here. Photosynthetic efficiencies can be
boosted significantly if one can engineer organisms that selectively
capture and use photons of specific energies and also develop ways
to utilize the IR & UV portions of the spectrum.
As far as the "efficiencies" of small devices goes there is lots
of room for improvement. Nature has pushed along one set of vectors
and in at least one instance (the F0-F1 ATP Synthase enzyme) appears
to be very close to the limits. But most of the other enzymes that
Nature is based on aren't very close. Part of the problem is that
the energy currency (ATP) comes in only one value while the amount
of energy required for different operations may vary significantly.
It depends what you want to do with the energy. We know for example,
that you can compute almost for free. But if you actually want to
store a result (that erases bits) you are going to generate heat
and that limits how quickly you can do the computation without
expending further energy to remove the heat.
> To James, I say: Some laws were meant to be broken. I firmly believe
> that the laws of thermodynamics are not true laws of nature, but just
> a stop-gap milestone in our understanding of the physical world with
> which we interact. We will discover that even the laws of
> thermodynamics don't hold as laws at a certain level, and we'll make
> a great breakthrough that will totally change the way we manipulate
> our physical surroundings.
I doubt it. Asking for the removal of the laws of thermodynamics
is, IMO, a pipe dream. However, the most "limiting" factor we
seem to be dealing with is energy conversion efficiencies in
heat engines. There we have room for improvement from levels of
~30-35% to ~80-85%. But as multi-layer solar cells are improved
and combined into more complex structures it isn't clear how
good they might get (we are already past the ~30% limits many
once claimed). The efficies of electric to mechanical energy
conversion via electrical motors are already quite high but may
get even better if we can apply HTSC.
If you are willing to run your mind on a large reversible
computer that computes using only latent heat energy for power
and allows simple conduction or convection to remove the heat
produced from erasing bits, then you will be able to run very
very efficiently indeed.
> My only curiousity is if this breakthrough will occur in my lifetime.
> If the sentiment that you hold, that the laws of thermodynamics are
> indeed strong laws and are a limiting factor, is widespread amongst
> the scientific community, then sadly I say I doubt it WILL happen
> within my lifetime.
I believe the laws of thermodynamics to be very strong but there is
a wide range for improvement that you should see within your lifetime.
Think, for example, of biological systems where all of the enzymes
are redesigned to match the energy denomination used precisely to the
energy required to drive a reaction forward.
This archive was generated by hypermail 2b30 : Sat May 11 2002 - 17:44:25 MDT