Hello,

I submitted my Master thesis http://e-collection.library.ethz.ch/view/eth:6011 and would like to thank the Intel MKL team for their fantastic support during those 6 months.

The code is available for download from http://hpsfo.bitbucket.org/ and offers the fastest implementation for Submodular (discrete) function minimization also thanks to using Intel MKL, TBB and other Intel products e.g. VTune.

Best regards,

Giovanni Azua