Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: $Revision: 1.6.2.16 $; site ISM780.UUCP
Path: utzoo!watmath!clyde!bonnie!akgua!whuxlm!harpo!decvax!yale!ISM780!marv
From: marv@ISM780.UUCP
Newsgroups: net.arch
Subject: Re: Re: attocode?
Message-ID: <27700001@ISM780.UUCP>
Date: Mon, 24-Jun-85 11:39:00 EDT
Article-I.D.: ISM780.27700001
Posted: Mon Jun 24 11:39:00 1985
Date-Received: Thu, 27-Jun-85 04:47:15 EDT
References: <2304@sun.UUCP>
Lines: 26
Nf-ID: #R:sun:-230400:ISM780:27700001:000:1225
Nf-From: ISM780!marv    Jun 24 11:39:00 1985


>> How did we manage to skip millicode?  Anybody know?

>Did you ever hear of a milliprocessor?

No.  However Standard Computer did build a family of machines (circa 1964),
the IC400, IC600.  Each machine had two alu's.  One was 12 bits wide and one
was 38 bits wide.  The 12 bit alu was used for doing address arithmetic for
the program in the control store, the 38 bit alu was used for performing
operations of the target machine.  The 12 bit alu was called the mini-engine,
while the 36 bit one was called the main-engine.  The program in the control
store was refered to as Miniflow*.

The Machine was used to emulate the IBM 704, 709, 7094, 7074, and 1401 as
well as some machines designed by me (e.g. the EX01 and ES01) The EX01 had
a dynamically alterable control store.  A programmer using the ICAP assembly
language could define a "MINI".  This was anagoulous to a ordinary macro
except that the body of the MINI was a "miniflow" sequence to implement a user
defined target instruction.  The Object module produced by ICAP containd code
that was loaded control storage as well as code that was loaded into main
storage.

* Miniflow is a trademark of Standard Computer Co.

			Marv Rubinstein,
			Interactive Systems