**4. Results and Discussion**

In this section we report the results of tests designed to characterise the performance of the SpiNNaker system running a pattern matching algorithm implemented with the MPI programming paradigm.

As a preliminary evaluation, we measured the execution time and memory usage of the MPI primitives implemented on SpiNNaker that we will use for implementing the parallel FED algorithm. MPI\_BARRIER, MPI\_SEND/MPI\_RECV, and MPI\_BROADCAST primitives will be tested and the results are reported in Section 4.1. Next, we performed an evaluation of the performance of FED algorithm implemented with MPI and executed on the SpiNNaker and on the CPU-based architectures. This last analysis aims to evaluate the scalability and power efficiency of the SpiNNaker platform when compared with a standard architecture. The results are detailed in Section 4.2.
