From: Eugene.Leitl@lrz.uni-muenchen.de
Date: Fri Dec 29 2000 - 02:26:51 PST
Adam Rifkin wrote:
> becomes cutthroat. Because once we get up in the Exabytes range
> (what'll that take, 10 years?) -- that's it, it's more storage than any
> person can meaningfully use.
Rotated bits are surface based. You can add tricks like more physical
platters or virtual (layers you can focus on), and use tricks like
spectral hole burning to pack more bits into the same volume, but
surface-based storage remains limited by the surface, and the minimal
pattern size. And even with manipulative proximal probe you can't poke
holes smaller than atomar into your substrate. (In practice your holes
will be many atom diameters across). Kazing, you suddenly ran out of
surface. Linear log plot suddenly goes horizontal. Unless, and just-in-time:
Assuming you can pack a bit into a cubic micron of macromolecular
crystal (conservatively, and shrinking feature size by one tenth only
gives you another three orders of magnitude), that's not so awfully many
bits in a cubic cm. There's plenty of room at the bottom, but after you've
reached the bottom the only thing left is scaling up volume. Which has
to be cooled, and interconnected. Which, rather jarringly, ends the linear
log plot. (But not in 2010, oh no).
As to storage, how much high-res multimedia coverage of 80 years of human
life is going to take? And since this is storage which can also compute on
essentially the same footprint, how much molecular circuit volume do you
need to make a machine as smart as yourself? I'd think no smaller than an
orange, perhaps a walnut.
And if these things are persons, don't you think they'd object to your
use of "meaningfully"? With the same logic you can say no one needs as many
atoms as there are in the human body, and those trapped in cubic miles of
life support around it. While there is some truth to the claim of a lack of
utilization efficiency, I'm sure many people see it in a somewhat different
way.
This archive was generated by hypermail 2b29 : Fri Dec 29 2000 - 02:34:51 PST