Re: How does Mma work on a Mac Powerbook 170?
- To: mathgroup at yoda.physics.unc.edu
- Subject: Re: How does Mma work on a Mac Powerbook 170?
- From: mek at guinan.psu.edu (Mark E. Kotanchek)
- Date: Mon, 21 Sep 92 11:45:38 EDT
Hi Folks,
Thanks to Magnus, Stefano, Tin Yau, Sherman, Elliot, Bill, David, and Ed for
their comments on the topic. The general consensus seems to be that the NeXT
advantages of a "real" operating system and faster processor makes Mma much
better there than on the Mac--expecially the Powerbook which has that
hardwired limit of 8 MB of RAM.
I took Ed's timing numbers and ran the same commands on my NeXTstation (25
MHz '040 with 32 MB of RAM and we got:
Timing[N[Pi,500];]
Mac SE/30 -- 1.58 secs
NeXTstation -- 0.15 secs (Mac is 10.5 times slower)
Timing[Inverse[Table[Random[],{100},{100}]];]
Mac SE/30 -- 39.9 secs
NeXTstation -- 3.3 secs (Mac is 12.1 times slower)
Timing[10000!][[1]]
Mac SE/30 -- 220.7 secs
NeXTstation -- 59.6 secs (Mac is 3.7 times slower)
So it would appear that for numerical stuff, the '040 is a significant
advantage. I'm not sure that this advantage would hold up for symbolic
computations; however, the need for RAM by symbolic computations coupled
with the relatively primitive Mac OS probably implies that I'd run into
problems on the Mac once I got into anything really interesting. In summary,
Mma on the Powerbook would be functionally acceptable but not necessarily as
pleasurable as on the NeXT.
The bottom line is that I'll hang onto my pennies and eventually spring for
the NeXT--which has a much better user environment for day-to-day use as
well as better Mma performance.
Thanks for the input,
Mark.
---
Mark Kotanchek
Guidance & Control Dept - 363 ASB
Applied Research Lab/Penn State
P.O. Box 30
State College, PA 16804
e-mail: mek at guinan.psu.edu (NeXTmail)
TEL: (814)863-0682
FAX: (814)863-7843