Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!husc6!mit-eddie!ll-xn!ames!sdcsvax!ucsdhub!hp-sdd!hplabs!hpda!hpesoc1!nicholso
From: nicholso@hpesoc1.HP.COM (Ron Nicholson)
Newsgroups: comp.arch
Subject: request: Old Iron performance
Message-ID: <5030005@hpesoc1.HP.COM>
Date: Wed, 30-Sep-87 19:41:37 EDT
Article-I.D.: hpesoc1.5030005
Posted: Wed Sep 30 19:41:37 1987
Date-Received: Mon, 5-Oct-87 01:38:14 EDT
Organization: Hewlett Packard, Cupertino
Lines: 31

Does any one out there have any performance information an old
computers; or references to publications with such information?

I'm trying to gather some relative performance information on
older and historic machines.  Because of the differences in the 
kinds of programs run then and now, a ball park figure is all
that I expect to find.

Now, it's unlikely that anyone has run Drystones on an IBM 704 or
Univac I.  So what I'm looking for is *any* early performance
measures.  So far, all I've seen are references that give processor
cycle times, memory sizes and such which are hard to relate to
comparitive performance. 

Perhaps if someone had ran some numerical program on a 6600, or some
such, that they now run on a PC this could provide a direct performance
measure.

Or if someone ported a program from historical machine A to a more
recent machine and noted that it ran X% faster.  If similar programs
run Y% faster on the manufacturer current machines and some
sythentic benchmark puts this machine at Z% of an 11/780, then we
might at least have an order of magnitude idea that the historical
machine is about 1/(X*Y*Z) mips for that class of programs.

I'll post a summary if requested, unless, of course, I don't receive
any information to summarize.

				Ronald H. Nicholson, Jr.
				Hewlett Packard
				...ucbvax!hplabs!hpda!nicholso