PDF
format.
This work presents the results of the Abingdon Cross Benchmark on the two PAPRICA systems, which are here briefly described focusing especially on their different architectural solutions. Starting from some considerations on their computational paradigm and hardware characteristics, the most efficient algorithm for the Abingdon Cross benchmark is discussed. The algorithm has been implemented on the two systems taking advantage of specific {\em ad-hoc} optimizations; the final {\em Quality Factor} is derived for both architectures and compared against that of other different non-commercial systems developed worldwide.