matrix operation in Octave

[Akos Marton]

Hi List,

We are working on a school-project which is about calculating product of large 
matrices. This project is trying to implement this operation on a heterogen 
cluster-architecture (e.g. multi-core Intel and Cell CPUs). Communication is 
based on our own communication method, later we wish to compare this network 
protocol to other already implemented and well known ones.

We would like to use Octave interface over our multiplication, because it is a 
popular and perfectly workable interface.

We would like to get information from Octave -,maintainers ,-developers: how to 
get the Octave's programming-interface and/or shell-interface to use together 
with our implementation.

This implementation is aimed to reimplement the multiplier operator or method 
(-It is a question how it is implemented now... ) in Octave to use our parallel 
matrix multiplication method instead of the common, already implemented one. 
Perhaps some overloading is needed.

I have some vague ideas, but I haven't been involved in Octave's source-code 
yet. Actually don't have much idea how to start with Octave developing, so I 
would appreciate some information on how to connect our project to Octave.

As a further goal we would like to get SVD to work together with our 
implementation.

Thanks in advance for your help!
Regards,
Akos Marton