🚀 go-pugleaf

In article <l3aq66INNlep@jethro.Corp.Sun.COM>, jdr@starflight.Corp.Sun.COM (Jon Roland) writes:

> Most of the progress that has occurred during the last century is the
> result of improving the efficiency of conversion of energy from one
> form into another by a factor of about 50, from less than 1 percent to
> about 90 percent of the thermodynamic limits. Further improvements
> are possible, and will be made, but not by another factor of 50.

Some of the progress, but not most.  I would say that most of the
progress of the last century is a result not of improving the
efficiency of energy conversion, but rather a result of improving
the efficiency of energy utilization.  We do far more with less
energy than our forebears, regardless of the efficiency of the
conversion process.  The difference is more in the application
of the energy than in its conversion.  Our trains are more 
efficient, but the far bigger improvement is that we no longer
need to travel physically between distant locations to conduct
all of our business.  Mail is transported more efficiently than
it once was, but facsimile transmission makes a lot of the mail
traffic unnecessary.  The challenge is to do more with less energy,
not to do a little more with the same energy. 

> Similarly, there are limits in computational power. The ultimate
> computers will involve the use of single electrons as message
> carriers and single molecular nodes as switching elements. That is
> an increase in power per unit volume of about 10^9, making possible
> about 10^18 NAND-gate equivalents per cubic centimeter. A lot of
> power, admittedly, but I am working on some problems that could bring
> even that kind of computing power to its knees.

Yours is an ironic choice of example.  A century ago you would not
have found this problem intelligible, much less significant.  What
does that suggest to you?

It suggests to me that a century hence this problem will, if even
remembered in a footnote of some history of technology, seem very
quaint indeed.  Our descendents will think us every bit as
technologically naive as we think our ancestors.  And they will
be right, just as we are.

In any case, as always, the solution to this problem when, not
if, it is found, will not be a matter of tweaking the physics of
the computer or the algorithm that you follow.  It will be a matter
of looking at the problem in a way that you and I cannot foresee.
It will be a matter of not needing to solve this particular
problem, rather than of solving it faster.  Your view is far too
narrow and limited.

One of Gordon Bell's remarks seems apposite:
"The cheapest, fastest, and most reliable components of a computer
system are those that aren't there."
--
----
Mark E. Slagle                                 PO Box 61059
slagle@lmsc.lockheed.com                       Sunnyvale, CA   94088
408-756-0895                                   USA