Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.2 9/18/84; site aurora.UUCP Path: utzoo!watmath!clyde!cbosgd!cbdkc1!desoto!packard!edsel!bentley!hoxna!houxm!mtuxo!mtunh!mtung!mtunf!ariel!vax135!timeinc!phri!pesnta!hplabs!ames!aurora!al From: al@aurora.UUCP (Al Globus) Newsgroups: net.space Subject: Re: Shuttle Computers and STS power consumption Message-ID: <372@aurora.UUCP> Date: Wed, 26-Jun-85 23:03:39 EDT Article-I.D.: aurora.372 Posted: Wed Jun 26 23:03:39 1985 Date-Received: Tue, 2-Jul-85 06:28:29 EDT References: <2322@mordor.UUCP> Organization: NASA Ames Research Center, Mtn. View, CA Lines: 24 > > A recent posting about shuttle cooling reminded me of an > article I ran across in an old (1980) IBM document. The article > described the onboard computers as System/4Pi model AP-101 processors. > The AP-101 appears to be a 3 KIP machine with an intensely ugly > architecture. Main memory was described as 104KW (36 bit) of core > (really core!) with 400 ns access time. What startled me were the > physical characteristics. Each of the five "processors" appears to > consist of two boxes. The boxes weigh 58.9 pounds and draw 350 > Watts!!!!! This would imply a system total of almost 600 pounds and > 3500 Watts. Did they actually use this refugee from the Smithsonian > on the Shuttle? I'm almost positive the answer is yes, they do and will continue to do so for the forseeable future. If it's not that particular model, it is a close relative. Sickening isn't it? Software note: software development for launch is supposedly very close to the critical path and definitely must see major improvements to support 24 launches a year. More horrors: there are some 250 microprocessors on the shuttle with a variety of incompatable languages, operating systems, and development systems between them.