[12080] in Public-Access_Computer_Systems_Forum
Nanoseconds and milliseconds: Correcting Dr. Internet
daemon@ATHENA.MIT.EDU (Walt Crawford)
Thu Mar 19 20:01:45 1998
Date: Thu, 19 Mar 1998 10:56:40 -0600
From: Walt Crawford <BR.WCC@RLG.ORG>
To: PACS-L@LISTSERV.UH.EDU
Reply-To: Public-Access Computer Systems Forum <PACS-L@LISTSERV.UH.EDU>
----------------------------Original message----------------------------
I'm used to "Ask Dr. Internet" offering bizarre "information"
and opinions...but you'd think such an Exalted Expert could
at the very least get basic measurements right.
milli, micro, nano: not a hard concept
Milli: one-thousandth
Micro: one-millionth
Nano: one-billionth (in American English)
60ns. RAM is, *theoretically,* one *million* times
faster than 60ms. rotating storage, not one *thousand*
times faster. That is, to be sure, a wholly specious
comparison, since it's the total time required to move
a needed string that counts, not the per-byte access
rate.
(In fact, a recently-released "solid state hard disk,"
consisting of several hundred megabytes of RAM with
enough battery backup to save to an attached real hard
disk in event of shutdown or power failure, proves to
be something like 25% faster than a real 11ms. hard disk
in applications tests. Neither a million, nor a thousand,
nor a hundred: not even double.)
In the interests of numeracy (and don't get me started
about the "billion personal computers," 80% of which
run without any known operating system...)
-walt crawford, br.wcc@rlg.org
To: PACS-L@LISTSERV.UH.EDU
