Path: utzoo!utgpu!news-server.csri.toronto.edu!mailrus!cs.utexas.edu!usc!ucsd!ucbvax!gnh-starport.cts.com!whitewolf
From: whitewolf@gnh-starport.cts.com (Tae Song)
Newsgroups: comp.sys.apple2
Subject: Re: TransWarp III
Message-ID: <9005041608.AA19359@apple.com>
Date: 4 May 90 16:53:37 GMT
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: The Internet
Lines: 21
X-Unparsable-Date: Wed May  2 90 at 09:50:45 (EDT)

>WDC does not have "a team of engineers" -- they have Bill Mensch and his
>wife and daughter. The original 65816 mask was laid out BY HAND five
>years ago and they have NEVER tried using the state of the art gate array
>approach that Tony Fadell had the connections to gain access to.

The 65816 was done by Bill Mensch and his SISTER.  Bill Mensch did the design
and his sister the lay out.  All the design was done by Bill's hand (to mylar),
because 1) it added to cost of making the chip and 2) he doesn't believe in CAD
programs for some reason.  His sister did the layout as was mentioned in a rare
interview several years ago in Compute magazine when the GS came out.  Oh
correction I ment it saved money by doing it by hand.

On CPU comparisons... the 68000-68030 buffered the clock and produce an
internal clock... 1/2 the external clock speed.  A 50Mhz 68030 is actually
running 25Mhz internally.  The 68040 fame is that it doesn't do this, thus a
25Mhz is twice as fast 68030, straight out.

The 65816 could've been design to be faster, but been design by hand some logic
circuits needed was sacrificed.  There's only 3 registers vs 16+ on the 680x0
and Ix86s.  There's probablely other things as well...
s.