If I remember correctly, Maple is capable of doing a lot of "memoization" & "hash consing", so that redundant computations are recognized & not performed redundantly.
For some computations this might help a lot.
At 05:53 AM 6/14/2012, Soegtrop, Michael wrote:
>Dear Dmitry,
>
>I compared some symbolic integration cases between Mathematica 8.0.4 and wxMaxima 5.27.0 and found that in some cases wxMaxima is quite a bit faster. I don't think that either Maxima or Mathematica or Maple is especially fast or slow in general, I think they just do some things in different ways, and a factor of 100 is not much if different algorithms are used, or e.g. simplifications are run on every step or only in the end. A single use case doesn't really tell a lot on which algorithm is better.
>
>But I think it would be interesting to collect a set of profiling cases, which are translated to Maxima, Maple and Mathematica and have a small database. I can do some Mathematica ports and tests.
>
>Best regards,
>
>Michael
>
>-----Original Message-----
>From: maxima-bounces at math.utexas.edu [mailto:maxima-bounces at math.utexas.edu] On Behalf Of Dmitry Shkirmanov
>Sent: Thursday, June 14, 2012 2:15 PM
>To: maxima at math.utexas.edu
>Subject: maxima is 100 times slower than maple
>
>Hello, list. I am comparing the speed of execution of maple and maxima programs. And maple is 100 times faster then maxima for symbolic computation. For example, to run the attached program maxima needs about
>50 seconds on my laptop. To run the same program(of course, syntax of program must be accordingly changed) maple needs about 0.5 seconds.
>I am using debian stable, maxima is configured to use sbcl, version of maxima is 5.27.0 Is there any possibility to speed up maxima?