Can nanotech make routers obsolete?
John Doe
Switches and routers will eventually become obsolete because of bottleneck
of the optical to electrical conversion.
http://gruia.blogware.com/blog/_archives/2004/8/20/127361.html
How will players like CSCO, EXTR, FDRY respond to the above threat?
Fred Chen
"John Doe" <do_not_...@nospam.org> wrote in message
news:<cg80i...@enews2.newsguy.com>...
Is it possible to avoid the speed limits of electronics in the
switching process? It seems that the switching laser circuitry would
need to run just as fast as (or faster than) the required optical
switching rate.
Phillip Thorne
On 21 Aug 2004 "John Doe" <do_not_...@nospam.org> reported:
>Switches and routers will eventually become obsolete because of bottleneck
>of the optical to electrical conversion.
>http://gruia.blogware.com/blog/_archives/2004/8/20/127361.html
[Researchers at Canada's Carleton University (Wayne Wang, Connie
Kuang) and U Toronto (Ted Sargent) develop a polymer-buckyball
material that acts as an optical switch.]
Okay, so I took a CCNA course a year ago; I should be qualified to
address this...
A router is a device that stands between two networks and translates
between their local addressing schemes. Local addressing exists
because (a) address spaces are typically limited and (b) even if
they're not, it's convenient to have local control. They can also be
used to segment a network, producing (a) smaller broadcast domains;
and unlike switches or bridges, provide (b) flexible logical
addressing, and (c) better admin tools.
(Network devices have a hardware address ("MAC"), which is assigned by
the manufacturer from a block allocated by a larger standards body.
Logical addressing, such as IP or Novell, acts above that. 32-bit
IPv4 naming is limited to approx. one billion logical addresses; many
fewer because of the way they're allocated. Small separate networks
permit the reuse of addresses without ambiguity. 64-bit IPv6 provides
more addresses than anyone will ever need -- probably -- but it's
*still* convenient to have local control.)
Any solution to the electronic-optical bottleneck is a threat to
*electronic* routers, not *routers* per se.
>How will players like CSCO, EXTR, FDRY respond to the above threat?
I should imagine they will adopt the optical switching technology into
a future generation of products, and write new marketing copy
trumpeting the improved performance.
Naming by NASDAQ symbol? Cisco, Extreme Networks, Foundry Networks.
(Thanks, Google.) Odd -- I'd do that in a networking group, but here
in sci.nanotech it's likely to confuse readers.
/- Phillip Thorne ----------- The Non-Sequitur Express --------------------\
| org underbase ta thorne www.underbase.org It's the boundary |
| net comcast ta pethorne site, newsletter, blog conditions that |
\------------------------------------------------------- get you ----------/