The openQCD code can be accelerated on BlueGene/Q machines by making use of QPX instructions.
Moreover, use of specific MPI functions can speed up global communications (this can be a significant effect on machines different from BlueGene/Q, e.g. clusters with Infiniband).

A modified version of openQCD-1.2 with these optimizations included can be found in the following tar file:

For further information read the file