🚀 go-pugleaf

In article mrmarcel@eos.ncsu.edu (Mike Marcelais) writes:
>In article noring@netcom.com (Jon Noring) writes:

>>I currently have a 386-33 pc with a Cyrix math coprocessor.  What I'd
>>like to know is how much faster would a 486-50 or 486-66 be when
>>calculating fractals using Fractint 17.2 (and for both integer and
>>floating point modes).

>Here's some benchmark times for fractint 17.2
>System:  486/50dx, 256K cache, 4M memory, DR-Dos 6.0
>         Tseng-4000 SVGA with Local Bus
>
>All options left at default settings, except where noted.
>
>Mandelbrot set:     1024x768x256
>floating point off       6.97
>floating point on        8.85


For my 386-33, it took 13.29 seconds for the 1024x768x256 FPU off, and
24.99 seconds for the FPU on.


>Plasma file, 1024x768x256, seed 42
>    Old method    22.19
>    New method    19.34

It took 57.34 seconds for the old, 51.03 for the new.


>Lypanov type, 320x200x256
>    Order  0   3:16.46

It took 9:11.35 for this one.


>Julibrot type, all defaults
>   320x200x256   3:30.31

It took 6:41 for this one.



Conclusion, the 486-50 was *approximately* 2 to 3 times faster in calculating
these fractals.  I would expect the 486-66 to be 10 to 20% faster than the
486-50.

Jon Noring



-- 

Charter Member of the INFJ Club.

Now, if you're just dying to know what INFJ stands for, be brave, e-mail me,
and I'll send you some information.  It WILL be worth the inquiry, I think.

=============================================================================
| Jon Noring          | noring@netcom.com        | I VOTED FOR PEROT IN '92 |
| JKN International   | IP    : 192.100.81.100   | Support UNITED WE STAND! |
| 1312 Carlton Place  | Phone : (510) 294-8153   | "The dogs bark, but the  |
| Livermore, CA 94550 | V-Mail: (510) 417-4101   |  caravan moves on."      |
=============================================================================
Who are you?  Read alt.psychology.personality!  That's where the action is.