Path: utzoo!attcan!uunet!world!decwrl!wuarchive!cs.utexas.edu!rutgers!att!dptg!ulysses!andante!alice!andrew
From: andrew@alice.att.com (Andrew Hume)
Newsgroups: comp.arch
Subject: architecture independence
Message-ID: <11444@alice.att.com>
Date: 8 Oct 90 15:49:27 GMT
Organization: AT&T Bell Laboratories, Murray Hill NJ
Lines: 13


	i am gathering evidence to support claims that various
refinements of an algorithm are (mostly) architecture independent.
the evidence are timings of the algorithm implemented in C and
compiled/run on various machines which can loosely be described as
MIPS R3000, SPARC, 68020, 80386, VAX and a Cray XMP. i believe
this to be a reasonably representative sample on which to base my claims.
i would appreciate feedback from anyone who disagrees (the feedback
should include the architecture and why it is important). i will not be
looking at any IBM 360 class machines. in order to conserve electrons,
email to andrew@research.att.com (research!andrew) is preferred.

	andrew hume