Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!aramis.rutgers.edu!pacbell.com!djo
From: djo@pacbell.COM (Dan'l DanehyOakes)
Newsgroups: sci.nanotech
Subject: Re: Dangers of Nanotech
Message-ID: <8905180515.AA01617@athos.rutgers.edu>
Date: 17 May 89 17:30:07 GMT
Sender: nanotech@aramis.rutgers.edu
Organization: Pacific * Bell, San Ramon, CA
Lines: 21
Approved: nanotech@aramis.rutgers.edu


In article <8905160500.AA05900@athos.rutgers.edu> Joseph.Bates@A.GP.CS.CMU.EDU writes:
>JoSH recently mentioned that in 10 to 20 years we will see an
>improvment of 1000 to 1 million in computing technology...
>
>Examples.  In 1965 the Linc, a high end "personal workstation" of the
>time, had 3K bytes and an 8us cycle time.  Today a RISC has about 10Mb
>and 40ns.  This is x3000 and x200, or about x30 and x10 per decade.
>
>So perhaps O(30) per decade is a better choice...

Hoom.  Seems to me that an increase in cycle time *or* an increase in
main memory will be an effective net increase in performance.  Then I'd
say that a x30 in memory and a x10 in cycle time is O(300)/decade, which
is still a bit lower than O(1000)...but do they heterodyne?



Ewige blumenkraft!

The Nearly Neurotic Net.Roach