Additionally, in the AI Engine application, the event::start_profiling and event::read_profiling API calls are specified to configure the AI Engine to count the accumulated number of clock cycles. Because you know how much data is coming out of the AI Engine (OUT_DATA_WINSZ), and this API reports the number of clock cycles it takes to receive the output data, you can calculate the port throughput (bytes per second). The AI engine application does this calculation and prints to the console during simulation. The longer the data is simulated, the higher the throughput accuracy. Currently, the number of iterations is chosen as 20.
Open the build/aiesim.log file to view the following throughput calculations for two output PLIOs:
DLBF Output0: Throughput 6.56944e+08 Samples/s, i.e. 2.62777e+09 Bytes/s
ULBF Output0: Throughput 6.59284e+08 Samples/s, i.e. 2.63714e+09 Bytes/s