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