On 16 May 2010, at 20:08, Alois Schlögl wrote: > What is your point here ? Some people are not as good in coding as > others, the function this student wrote is still useful. What does it > say about the performance of Octave vs. M ? Zero, Nil, Nothing.
Matlab has a feature called JIT (short for Just In Time compiler), which is not implemented in Octave (yet?). The JIT allows ML to some times automatically speed up code that is written in a way that is not the most efficient. The code snippets that Jaroslav highlited could be rewritten in a form that is much more efficient both in ML and in Octave. Usually if this optimization is done, the speed difference between ML and Octave is reduced a lot (and a few times even reversed). I think Jaroslav wanted to say that interesting benchmarks that are intended to assess the efficiency of Octave vs ML internals should be optimized as much as possible so to push both to the limit, otherwise what is actually being benchmarked is the efficiency of the JIT and not the maximum number-crunching power of the two programs. > BTW, this function is not part of the biosig benchmark. If the biosig benchmark is more optimized, then, as I said before, it would be very interesting to see the results of the same tests run on the same HW that you used for your benchmarking paper but with a newer version of Octave, as a lot of work has done since version 2.9.4+ to speed up basic Octave operations (mostly by Jaroslav himself) I am sorry I can not do this test myself as I do not have the same kind of hardware you used for your tests, nor I have an Octave 2.9.4+ version installed on my machines > Alois c. ------------------------------------------------------------------------------ _______________________________________________ Octave-dev mailing list Octave-dev@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/octave-dev
