[clean-list] Matrix-Matrix-Multiplication and Cleans memory management

[Siegfried Gonzi]

John van Groningen wrote:

> I have written a matrix multiplication that uses tranpose once.
> On my PowerMac G4 this version is 1.5 times as fast as the matrix
> multiplication in CLAS 0.7 (about 1 minute instead of 1.5 minutes).
> It also allocates less memory and probably uses less memory.
> I have included this program at the end of this e-mail message.
>

Thank you very much. Your code now tooks on my Performa 490sec (+-10sec due to
fluctuations) and the code now tooks on the Powerbook 510sec (+-10sec). And 15 to
20sec from the 510sec are used for garbage-collection.


By the way Omicron Basic tooks for the same matrix multiplication 1900sec. This
was really a surprise because Omicron Basic itself is written, for a great part,
in Assembler and produces PowerPC native stand alone code. And the default
build-in matrix-matrix multiplication is tuned.


Regards,
Siegfried Gonzi
[I would really encourage people to put numerical-code with a copyright remark on
the Clean homepage. I could not say it enough: that is the reason why C is in
academic areas so distributed: you can get for everything code]