Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watnot!watmath!clyde!rutgers!lll-lcc!styx!ames!pioneer!eugene
From: eugene@pioneer.UUCP
Newsgroups: comp.arch
Subject: Re: new Gould NPL; Whetstone MIPS (Back to 64 vs 32 bit)
Message-ID: <1197@ames.UUCP>
Date: Mon, 6-Apr-87 12:27:15 EST
Article-I.D.: ames.1197
Posted: Mon Apr  6 12:27:15 1987
Date-Received: Tue, 7-Apr-87 05:41:38 EST
References: <501@sw1e.UUCP> <1805@pyramid.UUCP> <326@elxsi.UUCP> <255@winchester.mips.UUCP> <332@elxsi.UUCP>
Sender: usenet@ames.UUCP
Reply-To: eugene@pioneer.UUCP (Eugene Miya N.)
Organization: NASA Ames Research Center, Moffett Field, Calif.
Lines: 36
Keywords: gould supermini, elxsi

In article <332@elxsi.UUCP> sherm@elxsi.UUCP (Michael Sherman) writes:
>In article <255@winchester.mips.UUCP> mash@winchester.UUCP (John Mashey) writes:
>>What are Whetstone MIPS? ....
>>	a) Are they DP MEgaWhetstones? (or SP?)
>>	b) Are they X times a VAX 11/780, if so, 
>>		b1) VAX/VMS
>>		b2) 4.3BSD?
>
>Sorry for not being more specific.  The numbers I mentioned were single
>precision megaWhetstones.  ELXSI's double precision numbers aren't much
>different since all registers are 64 bits.
>Michael Sherman

This brings up some things I have been thinking about regarding fair
comparisons.  How would you compare apples and oranges? (retorically,
send mail, not follow ups, I am summarizing on this one).
How would you compare say a 32 bit to a 64 bit to a 36 bit
architecture?  It's real easy to say, "If an application only needs
32-bit precision (or what ever architectural property, not just floating
point), then 64-bit (or whatever) is just icing on the cake."  Is it
more fair to do as dongarra and just say "everything in 64-bit" which is
the numerical analysts solution (no one does serious floating point in
32-bit, you only get 3-4 base-10 digits): so a "standard" minimum
is one way.  Another way seems to be normalization (I won't get into
this, it's the gold-plated VAX syndrome, and this is dumb [be glad these
people don't work for the NBS]).  What other means of making things fair
can you see?

From the Rock of Ages Home for Retired Hackers:

--eugene miya
  NASA Ames Research Center
  eugene@ames-aurora.ARPA
  "You trust the `reply' command with all those different mailers out there?"
  "Send mail, avoid follow-ups.  If enough, I'll summarize."
  {hplabs,hao,ihnp4,decwrl,allegra,tektronix,menlo70}!ames!aurora!eugene