From: James Higgins (jameshiggins@earthlink.net)
Date: Sat Jun 30 2001 - 13:39:10 MDT
At 11:29 PM 6/26/2001 -0400, Brian Atkins wrote:
>I asked on an AMD newsgroup and one person guessed 40 to 50ghz switching
>speed for .18 processes. So if the truth is somewhere between 40 to 100ghz
>I don't see how upping that to 210ghz in IBM's process translates to
>them claiming to be able to produce 100ghz clocked chips in 2003. hmm
Well, if you note the IBM statement applies to "communications"
hardware. I would imagine that an ASIC specifically designed for very high
speed routing could use significantly less transistors between latches (as
per Carl Feynman) than a general purpose processor such as an AMD
Athalon. Then again I'm merely attempting to extrapolate since I have very
little electronics knowledge...
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:36 MDT