Re: [Q] how fast is each machine with Mathematica ?
- To: mathgroup at smc.vnet.net
- Subject: [mg2673] Re: [Q] how fast is each machine with Mathematica ?
- From: wagner at bullwinkle.cs.Colorado.EDU (Dave Wagner)
- Date: Mon, 4 Dec 1995 01:41:33 -0500
- Organization: University of Colorado, Boulder
In article <49rklg$k50 at dragonfly.wri.com>,
Richard EASTHER <easther at cfi.waseda.ac.jp> wrote:
>In article <DIECwB.n3o at wri.com> smachov at ulc199.residence.gatech.edu (Scott J. Machovec) writes:
>>Kentaro Kondo (kkondo at flab.fujitsu.co.jp) wrote:
>>
>>: Is there information about Mathematica benchmark test ?
>>: ( for example, Timing[N[Sin[1/2],2500]][[1]] )
>>Pentium 90, 16 megs, Neptune chipset: 8.03 including loading kernel,
>>6.97 with the kernel loaded. Version 2.2 Enhanced in Win95.
>>
>
>Mac 7200/90 32 megs, (MacOS 7.5.2) Mathematica v 2.2
> - 4.68 seconds including loading kernal
> - 4.2 seconds with kernal loaded
Well, just for grins I fired up SoftWindows 2.0 on my Macintosh Quadra
610 with PowerPC upgrade card (50 MHz PPC 601, 25MHz bus):
- 99.5 seconds including loading kernel
- 87.8 seconds with kernel loaded
Well, I won't be throwing away my native PPC version of Mma just yet... :-)
BUT - Judging from the most recent benchmarks that I saw, the PowerMac
9500/132 is about 6 times faster than mine. Which means that one could
do this calculation on an emulated PC on a PowerMac in about twice the
time as for the 90 MHz Pentium. Which I think is an amazing testimonial
for SoftWindows (and the PPC chip).
Dave Wagner
Principia Consulting
(303) 786-8371
dbwagner at princon.com
http://www.princon.com/princon