Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!cwjcc!gatech!ncsuvx!ecemwl!jnh
From: jnh@ecemwl.ncsu.edu (Joseph N. Hall)
Newsgroups: comp.sys.mac
Subject: For you speed demons
Keywords: yow! fast!
Message-ID: <4039@ncsuvx.ncsu.edu>
Date: 25 Sep 89 22:19:22 GMT
References: <27395@shemp.CS.UCLA.EDU> <4360@internal.Apple.COM>
Sender: news@ncsuvx.ncsu.edu
Reply-To: jnh@ecemwl.UUCP (Joseph N. Hall)
Organization: North Carolina State University
Lines: 37

From this month's MIPS:

"A 25-MHz Am29000 brings powerful number crunching to Apple's NuBus
machines"

"...besting even a workstation like the DECstation 3100 in Dryhstones and
offering workstation-like performance in floating-point operations."

Currently there's only a "C" compiler for the board, and you have to compile,
link, etc., from a shell window.  You can run an executable file by
double-clicking its icon.

The card is the NuSuper coprocessor board from YARC Systems (5655
lindero Canyon Road Suite 721, Westlake Village, CA  91362  ph.
(818) 889-4388), and it runs about $6000 after you buy the card, compiler,
and associated goodies.

This caught my eye because of the persistent rumors that Apple is developing
an Am29000 machine, and is, I guess, an indication of the kind of number-
crunching capability such a "Mac" might have.  To wit:

		Drhy/sec	MFLOPS (sing)	(dbl) -- Linpack
YARC NuSuper	22230		2.166		1.416
Mac IIx		4060		.067		.076
DECstation 3100	15002		2.124		.930
Compaq 386/33	16000		.538		.426

The Mac IIx's miserable floating-point performance is a little TOO miserable,
and I think the reviewer's benchmark must have used the coprocessor indirectly
via SANE, which will indeed slow things down mightily.

Pleasant dreams!

v   v sssss|| joseph hall                      || 4116 Brewster Drive
 v v s   s || jnh@ecemwl.ncsu.edu (Internet)   || Raleigh, NC  27606
  v   sss  || SP Software/CAD Tool Developer, Mac Hacker and Keyboardist
-----------|| Disclaimer: NCSU may not share my views, but is welcome to.