More thoughts on building a Z-80 (64bit!!!)

From: Eric Smith <eric_at_brouhaha.com>
Date: Thu Oct 22 18:47:21 1998

> So what number were you using for the "raw material in the solar system?"
> And what manufacturing technology? Avogadro's Number is 6.02x10^23, a mole
> of carbon is only 12g. Using electron spin to store a bit as is being done
> today in quantum computers requires only 12.5 billon tons of carbon to
> store all of this information. That wouldn't even use up all of Mercury,
> much less the solar system.

I was using a rough estimate of the amount of silicon available in the solar
system, assuming that sufficient materials for dopants, oxide, metalization,
passivation, lead frames, bond wires, encapsulation, etc. were available as
well (probably not a valid assumption).

Quantum memory is not yet more than a laboratory curiousity. Even when it
becomes practical to store a bit using electron spin, it will still take
a large number of additional atoms to allow means of reading and writing the
bit.

I specifically said "Making this much memory using current technology..."
by which I meant 64 or 256 megabit silicon semiconductor DRAMs.

10.14x10^30 256 megabit DRAMs would be required.

One 256 megabit DRAM die currently uses about 3000 mm^3 of silicon, for a mass
of approximately 7 grams. So the total necessary mass of silicon would
be about 71x10^27 kg.

Approximately six times as much mass is used for the other materials that
comprise a packaged DRAM, so another 426x10^27 kg of appropriate materials
would be needed.

The mass of the solar system is somewhat more than 2x10^30 kg. However,
unless the gas giant planets have a higher than predicted silicon content,
the system contains less than 10^26 kg of silicon.
Received on Thu Oct 22 1998 - 18:47:21 BST