Newsgroups: comp.sys.atari.st
Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!sol.ctr.columbia.edu!cunixf.cc.columbia.edu!cunixb.cc.columbia.edu!es1
From: es1@cunixb.cc.columbia.edu (Ethan Solomita)
Subject: Re: Atari-To-Amiga Convert Info Source!
Message-ID: <1991Jun25.210655.18111@cunixf.cc.columbia.edu>
Sender: usenet@cunixf.cc.columbia.edu (The Network News)
Nntp-Posting-Host: cunixb.cc.columbia.edu
Reply-To: es1@cunixb.cc.columbia.edu (Ethan Solomita)
Organization: Columbia University
References: <1991Jun24.172215.816@colorado.edu> <1991Jun24.222634.30979@cunixf.cc.columbia.edu> <1991Jun25.173313.27320@colorado.edu>
Date: Tue, 25 Jun 1991 21:06:55 GMT

In article <1991Jun25.173313.27320@colorado.edu> chuj@horton.Colorado.EDU (CHU JEFFREY) writes:
>In article <1991Jun24.222634.30979@cunixf.cc.columbia.edu> es1@cunixa.cc.columbia.edu (Ethan Solomita) writes:
>>	The first point you make isn't a computer war, but a chip
>>war. The 040 is faster than the 486 at the same clock speed,
>
>I am trying to avoid any kind of war, this comp.sys.atari.st news is
>not for these types of postings.  SOOOO after this posting please
>send all responses to the posting to me and not post it here.
>
	I think that a discussion of chip technology is wholly
relevant to this newsgroup, as long as Motorola 680x0 chips are a
main part of the discussion. I shouldn't use the term war, I
consider this to be a "discussion".

>>(except under MS-DOS), but it is still close. The 040 25MHz does
>>appr. 18-22 MIPS depending on who you believe (I know how
>>meaningless that is, but we are talking all CISC chips, not RISC,
>>so isn't a terrible statistic). I believe the 486 is about 12.
>
>MFLOPS is a better measure than MIPS, the 68040 25 does about 3.5 MFLOPS,
>the i486-25 with Weitek 4167 does 10+ MFLOPS and the ranges of MIPS for
>the 486-25 is from 11 to 15 MIPS.
>
	I don't have MFLOPS statistics myself, so I'll take your
for truth. You mention a Weitek math coprocessor, which is not
made by Intel. That is something separate which is an additional
cost beyond the 486. Also, MFLOPS stands for "millions of
floating point operations per second". That is very important for
some programs (such as ray-tracers), but less relevant for other
programs, such as DTP. That's why benchmarks stink. For a DTP
application, MIPS is more important, although I'd prefer to hear
some Dhrystones and SpecMarks numbers as well.
	-- Ethan

FF buckets of bits on the bus,	FF buckets of bits.
Take one down,			Pass it to ground,
FE buckets of bits on the bus.

