Newsgroups: comp.sys.mac.hardware
Path: cantaloupe.srv.cs.cmu.edu!crabapple.srv.cs.cmu.edu!fs7.ece.cmu.edu!europa.eng.gtefsd.com!howland.reston.ans.net!wupost!crcnis1.unl.edu!moe.ksu.ksu.edu!hobbes.physics.uiowa.edu!news.iastate.edu!cmmiller
From: cmmiller@iastate.edu (C. M. Miller)
Subject: RESULTS of Mathematica Speed Tests!!
Message-ID: <C51I8u.2ux@news.iastate.edu>
Sender: news@news.iastate.edu (USENET News System)
Organization: Iowa State University, Ames, IA
Date: Tue, 6 Apr 1993 02:14:52 GMT
Lines: 84

Well, here are the results of the Mathematica test which I posted to
this newsgroup.  The "test" was the following command:

>Plot3D[((-2*9000)/(2*3.1416*((x-5000)^2+(y-8000)^2+
>81000000)^1.5))+((-3*9000)/(2*3.1416*((x-10000)^2+
>(y-1000)^2+81000000)^1.5))+((4*2000)/(2*3.1416*
>((x-7000)^2+(y-10000)^2+4000000)^1.5)),
>{x,-5500,19500},{y,-5500,19500},PlotPoints->50]


I was just curious how fast the plot command would be executed on
various Macintosh machines as well as other personal computers and
workstations.  The results are posted below:

Machine		System		Math vers.	# of trials	time, min

PB 170		7.0.0 with	2.1		2		2:08
		tuneup/8MB
		RAM/5MB for
		Mathematica

DEC 5000	Ultrix v4.2a	2.1 for		1		0:25
		DEC RISC	
					
IIsi		7.1/cache@96MB	1.2f33Enh.	1		4:30
		25MHz/5MB RAM/
		3MB for Math./
		w/ 68882

C650		7.1/8MB RAM			2		0:32

Q800		8MB/Cache@384/	1.2				1:01
		4MB for Math.	

Sparc		SunOS4.1.3					0:14
Station		40MB RAM

SGI Iris/4D	R3000 RISC					<0:01
		processor
		version

Sparc		SunOS4.1.2	2.1 				0:26
Station2

IIsi		7.1						3:15

NeXT		NeXTSTEP 2.1	1.2				2:38
Cube		68030 based/
		w/ coprocessor	

NeXT		NeXTSTEP 3.0	1.2		5(ave)		0:52
Cube		68040/25MHz/
		20 MB RAM

IIsi		17MB/8MB for	2.102 Enha			3:15
		Math.			
		w/ 68882

NeXT		16MB RAM/			1 		0:37
		25 MHz 040/
		Workspace
		Manager 2.1

Funny how the IIsi running at 25 MHz is slower than other equivalent
machines, lots slower in fact.  Perhaps the version of Mathematica
makes a difference or the fact that not much RAM was allocated.
Another interesting thing is how fast the SGI did it.  Wow. 

Basically, though, I wouldn't draw any conclusions from this data.  It
seems that Mathematica's speed is dependant on a lot of variables.  I
was just curious how different machines would measure up.  

Well, if you have any questions or if I forgot something, just drop me
a line at "cmmiller@iastate.edu".  

Chad

PS  If the spacing of the above table doesn't come out right on your
machine, tell me and I'll mail you a copy of this in a binhexed Word
5.1 document.  




