SCOOP running time penalty

Hello all,

gc for processors works well out of the box, BUT I have a running time penalty for longer inputs on a machine
with four cores hyperthreading in comparison to the sequential program.

I have set up a project homepage under http://pfhp.berlios.de (german)
(I will translate, e. g. "PFHP im Überblick" if necessary)

Likewise there is a GIT repository, it is a simple set up but the files should be accessible through:

$ git clone git://git.berlios.de/pfhp

Anybody a clue on parallel optimization? Yes, I read producer consumer howto, but I do not know how to
implement it myself and if it will be helpful!

Thanks in advance
Gerrit

------------------------------------