IIgs kills ARM-A2 prototype! Film at 11.

[toddpw@ugcs.caltech.edu (Todd P. Whitesel)]

Sources:
http://www.acorn.com/acorn/company/technology/history/technical.html
http://www.amazon.com/exec/obidos/ASIN/0201624109/002-5028283-3601025
http://www.uce.com/machist.html
http://www.hypermall.com/History/AH10.html
http://www.hypermall.com/History/AH11.html
http://falcon.cc.ukans.edu/~dangermo/IAUG/TECH/system.html
http://homepage.netspaceonline.com/~amigatec/library/amiga_history.html

The amazon.com book is out of print, but due to the Newton the 1994
Addison-Wesley printing should be readily available in libraries and used
technical book stores. It provided the clock speeds and some dates.

The first ARM1 silicon was manufactured in April 1985. The second generation
ARM2 was developed from 1985 to 1987 and shipped in the Archimedes in 1987;
this chip ran at 8 Mhz but the Archimedes was rated at 4 MIPS by Acorn. Math
hardware for the Archimedes appeared in 1989, along with the 25 Mhz ARM3.
The Archimedes 540 was launched in 1990, with an ARM3 on its motherboard and
built-in floating point hardware accessed via ARM coprocessor instructions.

On the Apple side, the Mac128's lifetime was January 1984 to April 1986.
The IIgs was developed from "late summer 1984" to September 1986. The first
Macintosh to support hardware floating point as an option was the Mac II,
introduced in March 1987.

And the Amiga: the first Amiga was introduced in July 1985 (68000 @ 7mhz).
The A2500 was introduced in 1987 (68030 @ 25 mhz). The A3000 was introduced
in 1990 (68030 @ 25 mhz plus hardware floating point).

(a) Note that the first ARM CPU's became available halfway through the
Mac128's product cycle, but very early in the IIgs' development cycle.
According to the Apple II history, the "Golden Gate" processor card projects
had been cancelled before the IIgs project was begun (the IIgs project was
even code named "phoenix" after the ashes of the "Golden Gate" project).

This raises two questions:

    1. Why would _anyone_ want to kill the Macintosh a year after its
	introduction in order to replace it with an ARM-A2 ?
    2. Why would _anyone_ within Apple be working on an ARM-A2 prototype
	six months after the IIgs project was started ?

Charlie, if you saw a prototype ARM-A2, it had no place in Apple's history.
I think that what you saw was not an ARM but some other processor.

(b) Note that the history of hardware floating point heavily favored the
_Mac_, with the Amiga and Archimedes both waiting until 1989/1990. I have
only cited machines which were "top of their line" at their introduction.

Supertimer, the Amiga vs. Archimedes contest you saw was rigged against the
Amiga somehow. Both machines had various speed options, so unless you can
remember a lot about them, we'll never know why the Archimedes did so well.

Todd Whitesel
toddpw @ ugcs.caltech.edu