Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watnot!watmath!clyde!rutgers!husc6!seismo!amdahl!chongo
From: chongo@amdahl.UUCP
Newsgroups: comp.arch,comp.sys.nsc.32k
Subject: Re: NS32532
Message-ID: <6166@amdahl.UUCP>
Date: Sat, 11-Apr-87 04:46:13 EST
Article-I.D.: amdahl.6166
Posted: Sat Apr 11 04:46:13 1987
Date-Received: Sun, 12-Apr-87 03:26:02 EST
References: <4173@nsc.nsc.com> <143@unsvax.UUCP> <6145@amdahl.UUCP> <744@instable.UUCP>
Reply-To: chongo@amdahl.UUCP (Landon Curt Noll)
Organization: Amdahl Coup, UTS Products Hen house
Lines: 51
Keywords: New product announcement
Xref: utgpu comp.arch:835 comp.sys.nsc.32k:48

In article <744@instable.UUCP> amos@instable.UUCP (Amos Shapir) writes:
 >In article <6145@amdahl.UUCP>, chongo@amdahl.UUCP (Landon Curt Noll) writes:
 >> 	What was their claimed 'simulated' Dhrystones of a 32332?
 >Please support this claim by numbers: 

OK.  Let is take the 'demo claim' made my NSC during the stockholders
meeting of 1985.  They 'said' that the 32332 = 3*32032 at the same clock 
rate.  Now let us take a later claim made by a NSC person on the net: 

	>From: nsc!roger
	>I don't care which compilers you use on which version of the
	>benchmark, the 532 is better than 5X improvement over 
	>todays 332 and over 11.7X faster than the 32032.

Now if we note that:  

	(532_rate/332_rate) / (532_rate/032_rate) == 332_rate/032_rate

his numbers show that: 32332 = (2.34)*32032.  In the very same article, 
the same person writes:

	>From: nsc!roger
	>the 32532 is 2.53X the 32332 at the same frequency
	>the 32532 is 3.86  the 32032 at the same frequency

This set of figures show that: 32332 = (1.53)*32032  (rounded in NSCs favor)

Returning to the question at hand.  The 'benchmark factor' could be shown
to be: 3/2.34 or 3/1.53.  If the 32532's 18000 Dhrystones underwent the 
same change it would be reduced to: 14040 or 9180.

<<i'll reply in AMOS's other questions in another article later>>

Let me state again:  I don't hate the 32000 chip set.  The system I now use
at home is a Symmetric 375 with a 32016 CPU.  (which I bought after
I left NSC)   I'm not 100% down on the 32000 chip set.

On the other hand:  I once had to select between a 32332 (provided to me
for free) and a 68020 (paid out my own pocket) for use in a data encryption
box.  Both ran at the same Mhz and used the same 0 wait state memory.  Both 
had an MMU.  Both took interrupts.  Both programs were hand tuned to perfection.
(I wanted a fast encryption box, not a benchmark!)  The 68020 won hands down.

On the third hand: My next home system will be a Sun, unless someone can show
me a similar priced 32000 based system that will out perform a SUN III.  I'm
open to suggestions.

chongo <> /\oo/\
-- 
[views above shouldn't be viewed as Amdahl views, or as views from Amdahl, or
 as Amdahl views views, or as views by Mr. Amdahl, or as views from his house]