Hot computers [was Re: NEWS: From Genes to Difference Engines]

From: Robert Bradbury (
Date: Sat Apr 15 2000 - 08:52:53 MDT

On Fri, 14 Apr 2000, Ken Clements wrote:

> A few years ago I remarked to Ralph Merkle "You keep saying that no one
> will ever really build a computer out of nano rod logic, but I suspect
> that at some point in the future a museum staff will do so to see if it
> would have worked."

I would tend to agree with Ralph, in that I think we will have things
like single electron transistors before we can build nano-rod-logic
diamondoid computers. However, titanium carbide or even better a
hafnium carbide rod-logic-computer would really crank. Why?
Highest melting points. Means you can put the CPU elements much
closer together due to decreased cooling channel requirements,
thus reducing the propagation delays resulting in greater throughput.
Interestingly enough, your coolants of choice would probably be
liquid beryllium if you wanted the lowest mass (lower pumping costs)
or liquid tungsten if you wanted the highest operating temperature.

Simulations would need to be done however to make sure the "rods"
still slide properly and don't become too soft or sticky. Since
you probably have to design in allowances for thermal expansion,
this might be a computer that only functioned accurately after you
warmed it up to the toasty temperature of 3000-4000 deg K.

Interestingly enough, there would clearly be uses for such computers
in space probes designed to descend through stellar atmospheres.


This archive was generated by hypermail 2b29 : Thu Jul 27 2000 - 14:09:19 MDT