Re: New technology around the corner [slashdot.org]

[Tim May <tcmay@got.net>]

At 10:32 AM -0800 2/18/98, Jim Burnes wrote:

>Does anyone else get the feeling that we are on the cusp of
>serious exponential change in technology?   I know its been
>going exponential for a while, but now I'm really starting
>to feel it.  Thats major breakthroughs in two weeks:
>
>1. 170 TB Polymer memory sandwiches (OptiCom)
>2. Flat Plastic Video Screens using Light Emitting Polymers "LEPs"
>(Cambridge)
>3. massive DRAMS (Hitachi)

No, I don't think we're on any kind of cusp of a growth curve. If anything,
several things are slowing down.

Beware the "press release." The items above are just a few of literally
thousands of such announcements.

Remember how "laser pantography" was going to revolutionize chip-making,
and even make "back yard fabs" possible? How about "silicon on sapphire"
and how it would obsolete Integrated Injection Logic? (You don't remember
I-squared L? Shame on you, as it was scheduled to put Intel out of business
by 1976).

Or how about plastic cubes that can store terabytes. Or wafer scale
integration. Or e-beam addressed memory. Or neural networks. Or fuzzy logic.

("Hey, Tim's being a negativist. My new rice cooker says it has fuzzy logic
in it.")

And then there's the whole universe of speculation about quantum computers,
DNA computers, nanotechnology, etc.

The fact is that R&D labs partly run on hype. And journalists are willing
to interview researchers to generate stories.

A particularly interesting place to read about all the Latest and Greatest
technologies destined to replace silicon is "Electronic Engineering Times."
EE Times has for at least 15 years been running breathless hype about such
technologies as neural networks, brain machines, organic semiconductors,
optical memories, and on and on.

(These articles are often interesting, too. I just don't get too excited by
some researcher's predictions.)

>ObCrypto: Since polymer transistors seem to be necessary to implement
>polymer memory, how does this effect computation speed?  The polymer
>transistors must be around 30-40nm and low power.  How long until we
>have Polymer PGA,PALS,CPUs?  How fast will they be?  Is it time to add
>a few more bits to the public key?  Are there any cipher attacks that
>would benefit from obscenely large amounts of fast memory?  Sorry if
>this is a naive question to the theorists.

If you literally mean "add a few more bits," as in 4 or 5 or so bits, that
pretty much applies every few years, by the usual Moore's Law sorts of
calculations. (Except of course it makes no logistic/administrative sense
to literally add a few more bits...better of course to pick a "safely
large" size. And of course the strength is usually more dependent on the
underlying symmetric cipher key size (e.g., IDEA).

As for "polymer memory," or even Hitachi's "fewer electrons in the cell"
research, believe it when you see it. And when you can buy it. And buy it
cheaply.

But I don't get overly excited by announcements of new developments like
this. Perhaps following the industry for 25 years has made me jaded.

--Tim May


Just Say No to "Big Brother Inside"
---------:---------:---------:---------:---------:---------:---------:----
Timothy C. May              | Crypto Anarchy: encryption, digital money,
ComSec 3DES:   408-728-0152 | anonymous networks, digital pseudonyms, zero
W.A.S.T.E.: Corralitos, CA  | knowledge, reputations, information markets,
Higher Power: 2^3,021,377   | black markets, collapse of governments.