Path: utzoo!mnetor!tmsoft!torsqnt!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!caen!news.cs.indiana.edu!noose.ecn.purdue.edu!chunnel.ecn.purdue.edu!thomaz
From: thomaz@chunnel.ecn.purdue.edu (Jose E Thomaz)
Newsgroups: comp.sys.ibm.pc.hardware
Subject: Re: Difference between 386/33 & 486/25 not counting fp
Summary: Some more data.....
Message-ID: <1991Apr12.092814.25978@noose.ecn.purdue.edu>
Date: 12 Apr 91 09:28:14 GMT
References: <1991Apr9.085749.4568@agate.berkeley.edu> <1164@gistdev.gist.com> <1991Apr12.073828.20663@agate.berkeley.edu>
Sender: root@noose.ecn.purdue.edu (ECN System Management)
Organization: Purdue University Engineering Computer Network
Lines: 71

In article <1991Apr12.073828.20663@agate.berkeley.edu> ilan343@violet.berkeley.edu (Geraldo Veiga) writes:
>In article <1164@gistdev.gist.com> flint@gistdev.gist.com (Flint Pellett) writes:
>>
>>(they publish results of 5 different tests in 4 different OS'es and
>>about 4 different price categories.  Performance ratios between 386
>>and 486 machines seem to follow the same ratios as the ones below.) 
>>
>>DOS 386 Systems:	Dhrystone	Price
>>Micro Express /33	15,870		$4,998
>>Laser Ditigal 386/33	15,750		$4,296
>>Arche Legacy 386/33	17,134		$8,665
>>Northgate 386/33	17,131		$8,919
>>
>>AST 486/33		34,192		$4,490 (upgrade board)
>>Club American Hawk III	35,923		$6,495
>>NCR PC486/MC 33		35,234		$14,995
>>
>>The Dhrystone numbers drop down to around 12K for the 386 machines
>>at 25 MHz and down to 26K for the 486's at 25 MHz.  Unless your
>>arithmetic is different than mine, that works out to about
>>twice as fast, not 50% faster.
>

 Just to add some more data...

 I have a gateway 2000 25 Mz at home with 4 Mb of memory and a CYRIX co
processo (25 Mz too...)

 My department just bought a Northgate 486 25 Mz with 4 Mb of memory.


 When I first benchmarked it with standard system information programs a la
 QAplus I was disapointed. All results were vary close to my gateway.. 
 Even standard dos programs like the graphical preview in word perfect were
 not that impressive.. same for autocad (10, not the 386 version)

 Now the whole impression vanished when I used 386 specific programs..
 NDP fortran compiled programs run in less than half (almost 1/3) of the
 time. And only yesterday I was running MATLAB 386 benchmark set and here are
 some of the numbers..

 for a reference, MATLAB uses IBM XT index  = 1

 then               PC / AT = 1.357 (MATLAB 286 of course...)
           386/387 (20 Mhz) = 14.347  (The benchmark does not says brand..)
   my gateway/Cyrix (25 Mz) = 22.971
   Northgate 486 (25 Mhz)   = 64.736!!
                              ^^^^^^^^ 
         Now That is  a LOT FASTER! than my 25 Mhz 386.. And it was
pretty visible the difference as the different routines were executed..

Some other numbers given for comparison by the bench routines were:

   SUN-3/FPA = 19.426
   Sun-4     = 35.4397
   Spark     = 65.519  (Probably spark 1, I don't know!)
  
Of course, these are numerically intensive benchmarks, so every one just
use judgement of how representative these are for their own purposes..
I for one, am sold on the 486 for my large 386, 32 bit applications... I
just wonder if OS/2 version 2.0, being partially 32 bit will benefit as
much.. Anybody out there would have any information?

Well, just my $.02...

Eduardo 


Purdue University, Civil Engineering Department  -