Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!apple!well!yoo
From: yoo@well.sf.ca.us (Young-Kyu Yoo)
Newsgroups: comp.sys.amiga
Subject: Re: Sorry Guys, There is NO WAY!
Message-ID: <21968@well.sf.ca.us>
Date: 5 Dec 90 06:28:27 GMT
References: <9012032130.AA114840@vttcf.cc.vt.edu>
Organization: Whole Earth 'Lectronic Link, Sausalito, CA
Lines: 34

fmcphers@VTTCF.CC.VT.EDU (Frank McPherson) writes:
>In a recent comp.sys.amiga.whatever article, a non-.sig-using entity from the
>"Whole Earth 'Lectronic Link" states::
>>An Amiga 3000ux will crank out 5 Mips.  The 040 NeXTs crank out 15-20 Mips

Don't have a sig yet.  But I have a name, its Young-Kyu Yoo.

>	You say that the 3000ux will "crank out 5 mips", then say the 040 NeXT
>will "crank out 15-20".  Figures courtesy of Motorola, you say.  I have no 
>reason to doubt you, so I won't say anything about that, except for being 
>interested in how such an answer was arrived at.

Don't ask me.  I let the big boys do the benchmarks.  At any rate, those are
Motorola figures.  And several Unix experts have done their own benchmarks
(check out the NeXT newsgroup for details) and they confirm that the 040 is
three to four times faster than the 030.

>	I do, however, take exception to the assertions concerning the 
>MFLOPS of both machines.  "nearly 3" and "will be lucky to approach 1" isn't
>precise enough for this arguement.  They suggest that you don't know for sure,
>you're just pulling out some numbers which sound nice.  Don't post something
>you can't quote with better references, especially if you're knocking a 
>computer in its' home newsgroup.  You're going to get flamed.  Now, if you'd
>post real numbers along with their sources, that wouldn't be quite so bad....

I didn't want to bore you with the details.  Motorola claims 2.8 MFLOPs for
their FPU (which is built into the main processor).  Home-grown benchmarks
done by several in the NeXT newsgroup get numbers from 2.5 to 3.3 MFLOPs.

I assume the math coprocessor on the Amiga 3000ux is Motorola's 68882.
68882 performs at less than 1 Mip.  If you have any info to contradict this,
please state it.

My question:  Why attack my numbers without numbers of your own?